marriage and health: his and hers

Psychological Bul le t in2001. Vol. 127. No. 4, 472-503

Copyright 2001 by the American Psychological Association, Inc.0033-2909/OI/S5.00 DOI: 10.I037//0033-2909.127.4.472

Marriage and Health: His and Hers

Janice K. Kiecolt-GlaserOhio State University College of Medicine

Tamara L. NewtonUniversity of Louisville

This review focuses on the pathway leading from the marital relationship to physical health. Evidencefrom 64 articles published in the past decade, particularly marital interaction studies, suggests that maritalfunctioning is consequential for health; negative dimensions of marital functioning have indirectinfluences on health outcomes through depression and health habits, and direct influences on cardiovas-cular, endocrine, immune, neurosensory, and other physiological mechanisms. Moreover, individualdifference variables such as trait hostility augment the impact of marital processes on biological systems.Emerging themes in the past decade include the importance of differentiating positive and negativedimensions of marital functioning, the explanatory power of behavioral data, and gender differences inthe pathways from the marital relationship to physiological functioning. Contemporary models of genderthat emphasize self-processes, traits, and roles furnish alternative perspectives on the differential costsand benefits of marriage for men's and women's health.

The health-enhancing properties of personal relationships havebeen repeatedly documented. Data from well-controlled epidemi-ological studies suggest that social isolation constitutes a majorrisk factor for morbidity and mortality, with statistical effect sizescomparable to those of such well-established health risk factors assmoking, blood pressure, blood lipids, obesity, and physical activ-ity (House, Landis, & Umberson, 1988). Marriage is the centralrelationship for a majority of adults, and morbidity and mortalityare reliably lower for the married than the unmarried across avariety of acute and chronic conditions, including such diversehealth threats as cancer, heart attacks, and surgery (Chandra,Szklo, Goldberg, & Tonascia, 1983; J. S. Goodwin, Hunt, Key, &Samet, 1987; Gordon & Rosenthal, 1995; House et al., 1988). Thetwo major hypotheses for these disparities are selection and pro-tection, that is, healthier individuals are more likely to marry andto stay married, and/or they have more material resources, lessstress, more social support, and less risky health habits than theirunmarried counterparts (Umberson, 1992).

Despite the fact that married people, on average, enjoy bettermental and physical health than the unmarried, marriage's protec-tive effects are notably stronger for men than women (Berkman &Breslow, 1983; Litwak & Messeri, 1989). In contrast to theirmarried counterparts, nonmarried women have 50% greater mor-tality, compared with 250% for men (C. E. Ross, Mirowsky, &Goldsteen, 1990). Gender differences in social control of health-

Work on this article was supported in part by National Institutes ofHealth Grants K02 MH01467, R37 MH42096, PO1 AG16321, and R01HL58528.

Correspondence concerning this article should be addressed to Janice K.Kiecolt-Glaser, Department of Psychiatry, Ohio State University Collegeof Medicine, 1670 Upham Drive, Columbus, Ohio 43210, or to Tamara L.Newton, Department of Psychological and Brain Sciences, University ofLouisville, 317 Life Sciences Building, Louisville, Kentucky 40292. Elec-tronic mail may be sent to [email protected] or to tlnewton®louisville.edu.

related behavior appear to be one operative factor in this mortalitydifferential, because women are more likely than men to attempt tocontrol others' health; thus, when marriage promotes better healthhabits, these effects are relatively larger for men than women(Umberson, 1992). In addition, social integration is inversely re-lated to negative health habits (Berkman & Breslow, 1983), andthere are also gender differences in this regard; women's supportnetworks often include close friends and relatives as confidantes,whereas men typically name their wives as their main source ofsupport and the only person in whom they confide personal prob-lems or difficulties (Phillipson, 1997). Accordingly, even thoughboth bereavement and divorce contribute to poorer health andincreased mortality, marital disruption appears to be more detri-mental for men than for women (House et al., 1988).

Although loss of a spouse can provoke adverse mental andphysical health changes, the simple presence of a spouse is notnecessarily protective; a troubled marriage is itself a prime sourceof stress, while simultaneously limiting the partner's ability to seeksupport in other relationships (Coyne & DeLongis, 1986). Trou-bled marriages are reliably associated with increased distress, andunmarried people are happier, on the average, than unhappilymarried people (Glenn & Weaver, 1981). In fact, both syndromaldepression and depressive symptoms are strongly associated withmarital discord (Beach, Fincham, & Katz, 1998; Fincham &Beach, 1999). Given the centrality of the relationship, it seemslikely that marital functioning would have consequences for phys-ical health as well. However, on the basis of the evidence availableat the time, Burman and Margolin (1992) concluded that althoughmarital variables affected health, the effects of marital relation-ships on health status were indirect and nonspecific. Evidencefrom 64 journal articles published since their seminal reviewfurnishes a reason for revisiting this area, particularly in view ofthe growth in one key aspect of this literature, marital interactionstudies; prior work in this domain was limited to four publicationsthat described autonomic activity in samples of 2 to 19 couples(Burman & Margolin, 1992). In contrast, in the past decade re-

472

MARRIAGE AND HEALTH 473

searchers have documented a broad array of autonomic, endocrine,and immune alterations as couples interact; these data provide acritical context for understanding the interplay between interper-sonal processes and biological states (Ewart, 1993).

Overview

Drawing from empirical research conducted during the pastdecade, the present review refines and amplifies particular aspectsof the more comprehensive biopsychosocial "blueprint" for mar-

riage and health proposed by Burman and Margolin (1992) in their

seminal review. Specifically, the present review focuses on phys-

iological pathways leading from the marital relationship to phys-

ical health outcomes, along with the roles of depression, health

habits, and trait hostility (see Figure 1).

Three interrelated themes form the conceptual backdrop of this

review. The first theme, discussed above, is a major impetus for

this review and concerns the possibility of direct and specific

linkages between marital functioning and physical health. Accord-

Variabtes and Pathways Moderated by Gender-Linked Traits,Setf-Repfesenfeifofls, and Roles

PsychiatricSymptomatologyand Syndromes(e.g., depression)

NegativeDimensions of

MaritalFunctioning Biological Systems

(e.g., cardiovascularendocrine, immune,

neurosensory,neurophysiological)

Health Habits(e.g., substance

use, eating habits)

PositiveDimensions of

MaritalFunctioning

IndividualDifferenceVariables

(e.g., hostility)

Figure 1. Conceptual framework for organizing associations among marital functioning, biological systems,and physical health. Note that although this figure summarizes data that form the focus of this review (i.e., thosethat pertain to the direct and indirect pathways leading from the marital relationship to physiological functioningand physical health), it also illustrates hypothesized pathways along with bidirectional associations that areoutside the scope of this article (see Burman & Margolin, 1992). Marital quality and marital interaction are thetwo aspects of marital functioning that have been examined with regard to physiological functioning andphysical health. Positive and negative dimensions of marital functioning are composed of communicationbehaviors, emotional states, and cognitions (e.g., attributions and expectations) that are activated within specificmarital interactions, along with spouses' evaluations of the overall marital relationship and of specific maritaldomains (e.g., satisfaction-dissatisfaction, companionship, equality in decision making). Following the path-ways that are the focus of the present review—those leading from marital functioning to physiologicalfunctioning and physical health—the figure illustrates direct links from negative and positive marital dimensionsto biological systems. In the domain of marital communication, direct links from negative behaviors tophysiological functioning are supported; direct links involving positive behaviors are less evident, perhapsbecause paradigms that might best demonstrate these links have not been routinely used. Indirect pathways fromthe marital relationship to biological systems are mediated by health habits and psychiatric symptomatology thatare affected by marital functioning and that in turn impact biological systems (Whisman, 1999). Also shown areindividual differences variables that affect marital functioning and its associations with biological systems.Finally, the shaded background represents a triad of gender-linked factors that influence behavior, cognition, andemotion in close relationships. It is proposed that this triad helps explain gender differences in pathways leadingfrom marital functioning to biological systems and physical health.

474 KIECOLT-GLASER AND NEWTON

ingly, considerable attention is devoted to evaluating studies ofphysiological responses to marital interaction, including a discus-sion of the potential significance of these physiological changesfor subsequent morbidity and mortality.

The second theme concerns gender differences. As discussedbelow, in studies of physiological correlates of marital interaction,gender differences are robust and salient. In contrast, in researchon marital functioning and physical health outcomes, gender hasnot received consistent and systematic attention. Accordingly, wereview the past decade of studies on marriage and physical healthoutcomes with an eye toward gender-related differences, highlight-ing emerging patterns that suggest areas for future research. Inaddition, we consider conceptual perspectives on gendered pat-terns of physiological functioning and health status in marriage.Drawing on models of gender-linked individual differences, oneperspective considers how women's relational traits and self-processes could conceivably render them more responsive thanmen, psychologically and physiologically, to the emotional tone ofmarital relationships (Cross & Madson, 1997b; Helgeson, 1994).Another perspective considers how gender differentials in stressexposure that occur within the context of marital roles (e.g.,responsibility for and participation in domestic chores) mightcontribute to certain pathways leading from marital functioning todeleterious health outcomes (Glass & Fujimoto, 1994). As illus-trated in Figure 1, it is proposed that these gender-linked factorshave a pervasive influence on all pathways connecting maritalfunctioning with physiological mechanisms that contribute tohealth outcomes.

The third theme of the present review concerns the stress-socialsupport hypothesis or, alternatively, the social strain-social sup-port hypothesis, the major explanatory framework proposed byBurman and Margolin (1992). This aspect of their model wasproposed in order to account for both the protective and deleterioushealth correlates of marriage. Consistent with this explanatorymodel, a growing literature has suggested that negative aspects ofsocial relationships are often independent of positive aspects(Rook, 1998) and are important independent predictors of psycho-logical and physical functioning (Bolger, Delongis, Kessler, &Schilling, 1989; Stansfeld, Bosma, Hemingway, & Marmot, 1998).Moreover, research emerging in the past decade permits the sup-position that these two aspects of marriage may merit independentassessment in order to best understand links between marriage andhealth, for example, evidence that socially supportive behaviors,assessed prospectively, are associated with marital outcomes(Pasch & Bradbury, 1998). Accordingly, throughout this review,we highlight findings that speak to the possible independent effectsof positive and negative aspects of marital relationships, and wereturn to this theme when we discuss methodological implications.

Scope and Organization of Review

The literature search made use of the ancestry approach afterMedline and PsycINFO were surveyed using the terms maritalinteraction, marital adjustment, marital quality, marital conflict,and marital satisfaction; it spans 1990 through December 1999,beginning where Burman and Margolin (1992) ended their litera-ture review. For inclusion, researchers must have reported data onsome dimension of physical health and/or physiological function.Studies are clustered based on whether the dependent measures

were (a) objective physical health status or physiological data, (b)self-reported physical health, (c) pain outcomes, or (d) physiolog-ical assessments collected during or following marital interaction.Because this review focuses on the pathway leading from themarital relationship to physiological functioning and physicalhealth, studies in which marital quality was the dependent variablewere excluded; thus, for example, the effects of alcohol use onmarital functioning are not addressed (Jacob & Leonard, 1988,1992). Similarly, although there are obvious mutual influences,studies that simply used an illness as a stressor, without attemptingto relate changes in marital functioning to physical health, were notconsidered, for example, the literature on caregiver burden(Kiecolt-Glaser, Glaser, Gravenstein, Malarkey, & Sheridan,1996). Studies that examined family functioning were not includedif the marital relationship was not assessed separately.

Studies examining the link between marital functioning andphysical health outcomes are considered first. This part of thereview is organized into three sections, according to the dependentmeasure: objective physical health status or physiological data,self-reported physical health, or pain outcomes. For each of thesesections, the corresponding table provides a comprehensive list ofall studies that were located in the literature review. The discussionfor each section is designed to highlight salient themes, rather thanbe all-inclusive. Afterward, we review marital interaction studiesthat include physiological assessments and then address the rolesof depression, trait hostility, and health habits in the pathwayleading from marital relationship factors to physiological function-ing and health outcomes. Finally, conceptual perspectives on gen-der are discussed. The review concludes with a summary ofmethodological recommendations for the next decade of researchon marriage and health.

Marital Studies With Objective Physical Health StatusMeasures or Physiological Data

A survey of Table 1 reveals that marital relationship factorshave health implications for diverse medical conditions. For ex-ample, two studies observed relationships between marital vari-ables and immunologically mediated disease outcomes. The firstof these addressed oral health, an arena in which the immunesystem plays a key role. Both men and women who reported lowmarital quality were more likely to have periodontal disease anddental caries than those who reported high marital quality (Mar-cenes & Sheiham, 1996).

The second study addressed rheumatoid arthritis, another con-dition in which the immune system plays a central role. Differ-ences in the marital relationship were used to predict the impact ofinterpersonal stressors in the women's network on rheumatoidarthritis disease activity (Zautra et al., 1998). Although both im-mune function and clinician's ratings changed during a week ofincreased interpersonal stress, women who reported either morepositive spousal interaction patterns or less spouse criticism ornegativity did not show as large an increase in disease activity.These benefits were not found among women who reported onlythat their husbands were highly supportive during disease flare-ups. As the authors noted, this is the first study to successfullypredict changes in rheumatoid arthritis disease states using a prioridecision rules; the study's notable strengths include its prospectivedesign and objective assessment of disease activity. In addition, by


differentiating positive spousal interaction patterns and generalspousal support expressed during symptom flares, two differentand modestly correlated aspects of positive marital quality, theinvestigators revealed an important specificity with regard tomarriage-disease linkages in rheumatoid arthritis. Because spe-cific positive interactions, but not support, were implicated indisease activity, the authors recommended that future studies dif-ferentiate and assess various components of marital quality.

Other work underscored the impact of marital distress on car-diovascular function. Data from one innovative study of marriedwomen suggest that recollection of conflict is sufficient to alterblood pressure (Carels, Sherwood, & Blumenthal, 1998). Womenwho reported lower marital satisfaction exhibited higher systolicblood pressure and heart rate responses during marital conflictrecall than women with high marital satisfaction, after controllingfor reactivity in two nonmarital tasks. In a related domain, lowerscores on the Cohesion scale of the Dyadic Adjustment Scale(DAS) were related to elevated nighttime blood pressure and 24-hrdiastolic blood pressure among a population of early hypertensivemen and women (Baker et al., 1999). Gender was included as acovariate, making it unclear whether these associations held formen and women separately. In addition, among the participantswho scored below the group mean on Cohesion, more spousalcontact was associated with elevated evening blood pressure.

The spouse's behavior also appears to influence behavioralsymptomatology in neurological disorders. In a longitudinal studyof spouse pairs in which one member of the couple had Alzhei-mer's disease (AD), the caregiver's expressed emotion (EE), de-fined as expressions of criticism and/or overinvolvement towardthe care recipient, was predictive of increased negative behaviors(symptom manifestations of the underlying AD neuropathophysi-ology) in the impaired spouse (Vitaliano, Young, Russo, Romano,& Magana-Amato, 1993). In related work, behavioral symptoms ofParkinson's disease were coded from a videotaped interaction ofthe patient and spouse as they discussed changes in their livessince the patient's diagnosis (Greene & Griffin, 1998); patients indistressed relationships blinked less frequently and had longerblink durations than their nondistressed counterparts. Becausespontaneous eye-blink rate and duration can be affected by dopa-mine availability and thus can serve as quantitative correlates ofneurophysiological status among individuals with Parkinson's dis-ease, the investigators interpreted these results as evidence forneurophysiological symptom exacerbation among patients withParkinson's disease in distressed marriages. Taken together, theresults of these two studies indicate that marital distress andspecific negative spousal behaviors contribute to the worsening ofbehavioral symptomatology in two different neurological disor-ders. Analyses were not conducted by gender in either study, andthus no conclusions can be drawn about the role of gender in theseresults.

Of the studies reviewed so far, the only one that tested genderdifferences observed similar associations between marital func-tioning and health status for men and women (Marcenes &Sheiham, 1996). However, two additional studies showed thatmarital functioning had a stronger impact on women than on men.One such study used a substantial follow-up period, rather than across-sectional design, in order to establish a stronger and lessambiguous basis for links between marital processes and thedevelopment of important health outcomes such as cancer or heart

disease (Hibbard & Pope, 1993). Using medical records fromwomen and men randomly selected from among members of alarge HMO, the investigators found that health data from a 15-yearfollow-up were related to social role measures.(Hibbard & Pope,1993). For women, companionship in marriage and equality indecision making were associated with a lower risk of death. Men'smorbidity and mortality were unrelated to marital role character-istics. Data from another large longitudinal study reflect a similargender disparity (Appelberg, Romanov, Heikkila, Honkasaol, &Koskenvuo, 1996). Women who reported that they had "consid-erable conflicts" with their husband and who also reported workconflicts had a 2.54-fold risk of physician-certified work disabilityrelated to a variety of health problems in the ensuing 6 years.Neither work nor marital conflict was a risk factor for men.

In contrast, in two other studies, aspects of the marital relation-ship had a larger impact on men than women. In the first of these,mood self-ratings made while a couple watched the videotape oftheir conflict session that had taken place several days earlier weremodestly correlated with husbands' (but not wives') autonomicresponses recorded previously during conflict (Levenson,Carstensen, & Gottman, 1994). These data stand in contrast to theevidence that simple recall of marital conflict can heighten car-diovascular responses in maritally distressed women (Carels et al.,1998). However, unlike the procedure in Carels et al., the moodratings and the physiological data were collected at different timepoints, making interpretation problematic.

Apparent gender differences in a second study may have beenan artifact of statistical power. Husbands who reported that theydisclosed to their wives were less likely to die and/or be rehospi-talized within the next year following a myocardial infarction(Helgeson, 1991). This association held after controlling for anempirically derived biomedical index that is highly predictive ofmedical prognosis following a heart attack. With regard to genderdifferences, the beneficial effects of disclosure to one's spousewere significant for men (n = 63) but not women, but the smallnumber of married women (n = 14) precluded firm conclusions.

Taken together, the studies in Table 1 suggest that maritaldimensions other than global marital quality may contribute tohealth outcomes, including spousal conflict and overinvolvement,inequality in decision making, disclosure, and companionship (Ap-pelberg et al., 1996; Helgeson, 1991; Hibbard & Pope, 1993;Vitaliano et al., 1993). Because few studies assessed multiplemarital factors, it is unclear whether other aspects of the maritalrelationship make unique contributions to health outcomes orwhether they reflect the effect of a more global, underlying maritalquality dimension. One study that did evaluate two different as-pects of positive marital quality revealed novel evidence for spec-ificity with regard to marriage-disease linkages in rheumatoidarthritis (Zautra et al., 1998), thus highlighting the potential im-portance of disaggregating marital quality.

At the same time, global measures of marital satisfaction pro-vide critical information, and many of the studies in Table 1 eitherdid not assess marital adjustment or provided no data from scaleswith known norms that would allow inferences about the range andvariability of marital happiness in the sample. Given the lowerparticipation of dissatisfied couples in marital studies (Bradbury &Kamey, 1993), truncation of range may lead to underestimates ofthe effect of marital unhappiness. Restriction of range is also an

H

goQ"3

f"£"3£>•^<j

O•g

13

^

S

§§

!S)

WC

Cc

'ccd>

JU<u&

;/}OJ

PcdOJ

H

cs.Q

Wi

Sia3

artic

ipan

t 3'

Cu

55

,_ o > •" .. - 2 "S" ° § -o S • S"^ g - S ' S S - 2

|cl 1 |01S|| ill ill^ I1 ? 11 P ° 1 ! g 1 1 1 1 11 1 ig s . s * s g>^ a o •§ j u s 3 .f 2 •§ §Ml 2 -s !£|!"§ 3 2J si g a - llilt f^ l l l^s|^So.^. f ||| 1 f | | « B i § § ; S g £ ; g s | . S 3 §

^ ^ j j ^ 1 i « ! s I < - 2 5 ' 5 . , e .SS"§ o .a o

S o ^ ^ S c c ^ ^ S ^ ( U S ' r " - > ^ ' r t E ! - )

o, g, 'g o - a u ' ^ " 7 ' i W c i 1 i ° o" '"° ° "^ ^ o ^ ' O c ; S ' u - U H ^ ' - ' ^ " ^ 2 c > s iO O ^J C -5 D ^ O ^- 2- -S" o ^™ J^ ^ •— r^ 03

•g u^JJ | |"|s | J?SJ= ' ' oS >'o.iis>EJ3 ^ u '

> "3 i_S - H § ^ M

ca "O ^ S (y u UH*n T^ ^ " C '"" i ^ _, ^ •• nri_ G <& ^ " C r t ^ " " ' ^ C . S J " S n ^ E• S ^ 1 ^ S o t : ^ p S ^ S , 7^ rî '~c -S" — a °a -£'§ ^ "« J" § > ^ W I U G

c^ " i § 8 " •o'S§0!' '>i§i'"".= i | | f ^.g ,0 » I - « 1 g | | 0

^ ^ T u c 1 u'T'u.s' ' B . > S g o ' 3 3 2 i l.3 S •- '? — M ° '5 '•= w •= '> | 3 'g " b§" s a s .a 1 " 'f ^ S g g>-g ° "§ | ^ g"

•5* ®** ' ei) u ~" 4J

« cT .S . . .2 •£ So aoî ^ *> -a o73 U <2 -g £ ^ « g ^

^ § - S S a o . I H^P

11 .Ml I sfli-s^^ 2*1-1 S î«l|• ^ *_^ c ^ (d ^_r c *s . ' S5ldrt c d O . S O , cd C f e r t — Sc c 5 ' ^ ^ G S "-3 ° w y.

3^3 S^ l d 0 3 f l / ? § ~ o ' & J

fl fill 1 11 111H- S i-J nJ CJ

— H 2V5 S ^ >O O 4J >

1 " § - 6 1 i «s.|l I ill" *!!! o=IB -a !*§>! '?&î |"l?1 S g 5 | g ^ ^ x a ^^ . 2 c - n ^ e a ^ . S e a o o — « — SE S b C13^T3 f f l ^ ^ O t - ' O c S" O M ui o »> °-"i_ •- H — •- ..

"O U C r- ^*^ " QJ . " ^ V î*n ilil tiPitf* « ° ^ S i " c l ^ - o " ^ ^ c - ^ 8 lc ^ p ^ ^ o S 5 ^ * 0 ^"^ 5 E *H •>-< E2 S t^ —

o" ccd CP •£

0 0^* m O

ro 5S ~"

1—1 , cd a.,' OJQ £ U

- E. §< |2 *• >^ c Uc" ^ cT ^> c"I 1 l| |g.s? 5 "S $$ " —x S ? j

i«•

IIll'

111--°

ts =- I"«3 ^^|

sis *11

. f a S - c L S - c ^ g "11 '2 g II 1 S I

«}11

g E S S^

Is"

il1*& -x

E

l-

IS!

ill

1 ?

IIIill

-afl

11s

ss

-« - H

!lo ca .i! _,- U O

I I IUI

§ § S §r~o. a) TJ no3 ooa .« o> >H S „ -2 II§ S 1 •= cj= =3 -S S -21-" a-sJ- ^ o-•§,,{.

J I i= -1&•« 4-. tl, WJ

§ "g ° < *

o

^jss;

S

,

o2H

Rel

evan

t fin

ding

s

a

11C

11

•&

I

esig

n an

d in

depe

nden

i

Q

1

g•oc *-re c^^ OJ

OO *r;•5 3

reo.'oC0,

1OO

03

Alth

ough

im

mun

e fu

nctio

n an

d cl

inic

ian'

sra

tings

sho

wed

sig

nific

ant c

hang

e du

ring

"B .?.•3 i•S g"o re

ISre °

"re S>• - '§'oo reo >-

11£ -sij:

_^

1 ^Q Cd

i 00

•§!2-a•S|1 s

<L> —u cuS" a8,c_

"O

1 °cd j-

3 "o

S3-.

•a InO COu S2

t!o ^^<J G

•aS iIf^ ex

CO "e <u||1 |•is -5

CD ."S^ 0

S1

_3•o1cd

'iE

DM

cd

B 2S =5oo T3

posi

tive

spou

sal

inte

ract

ion

patte

rns

and

less

spo

use

criti

cism

or

nega

tivity

did

not

(D

1S-

freq

uenc

y of

pos

itive

inte

ract

ions

show

inc

reas

es i

n di

seas

e ac

tivity

.Lo

wer

Coh

esio

n sc

ores

wer

e re

late

d to

high

er n

ight

time

BP

and

24-h

r D

BF.

For

O-03>-l

r am

bula

toi

.c

oJD

« «s a

ross

-sec

tiona

l, 24

-hr

dco

llect

ion;

Coh

esio

n

U

©

II ..e

5.1~ — ' tf.c c

e

m ^T3 'g

l|

o2

•wU,

C>re

CQ

low

Coh

esio

n pa

rtic

ipan

ts,

mor

e sp

ousa

lco

ntac

t w

as a

ssoc

iate

d w

ith e

leva

ted

onQ(D

J=

OC

<ug3

'f"ScdEu -a00 JJ2 BU 0.J5 S.

even

ing

BP.

No

Gen

der

X C

ohes

ion

anal

yses

.Pr

e- t

o po

sttr

eatm

ent

impr

ovem

ent

in s

perm

conc

entr

atio

n an

d m

arita

l di

stre

ss.

"re

1£•ocsi.2 i

•3 """

m c

once

ntr;

stre

ss r

ated

S.'"300

Cut

" g£ 5

aupl

es i

n C

BT

to i

mp

com

mun

icat

ion

skill

s

U

i1

'« c

"c i3'5 °!

t"0 S-r^

«aU

3re

,

oU.

1C 0-

U £C *•

* §y rvE—

C/5fX

1"o

com

pare

d w

ith t

wo

nonr

ando

miz

ed c

ontn

c13

re

*2•c1

1re

t^ -o

« i i

ts;

IBI

= in

terb

eat i

nter

val;

FPA

= f

inge

r pu

ls2;

D

BF

= di

asto

lic b

lood

pre

ssur

e; C

BT =

u 3

1 scU 73

IIaj S11 1

CQ !?

men

t Tes

t; 1

SBP

= s)

•1, |" ^3 o.

I|

bb ..

"> "-i _(U

1 c

"o -S(U ^

1 M

I I ^J ^Q 13< £

I Ii IS ".cj i:•a '-3"$ <

jj oO.T3X «J

II ^

UJ ||W.. 00

O

"c 1"3 t;^r re^ «re -^

^ "P

— ..S|

•S; D.^ G

*£ ^

obvious problem when single items are used as the basis forevaluation of marital adjustment.

With regard to gender, some studies in Table 1 suggested sexdifferences in links between marital functioning and health status,whereas others did not. Although comparisons were hampered bythe fact that a number of the studies did not report analyses bygender or included only one gender, it is notable that the two largelongitudinal studies—some of the strongest empirical evidencebecause of the predictive designs and sizable samples—showeddifferences in disability and mortality for women, but not men(Appelberg et al., 1996; Hibbard & Pope, 1993).

Finally, the studies in Table 1 indicate that marital relationshipfactors bear significant relationships with physiological and objec-tive health status indicators. The broad influence of marital factorsis revealed by the number of different biological systems impli-cated (e.g., immunological, cardiovascular, neurophysiological)and by the number of distinct points in the disease trajectory atwhich marital factors appear to play a prominent role, includingpossible etiology (Hibbard & Pope, 1993); symptom exacerbationin chronic, degenerative illnesses (Vitaliano et al., 1993; Zautra etal., 1998); and prognosis-recovery following a life-threateningmedical event (Helgeson, 1991).

Marital Studies With Self-Reported Health

Self-reported health took two forms in the studies listed inTable 2. In the first, global perceptions of health or frequencies ofnonspecific health complaints served as the dependent variable. Inthe second, data were collected in regard to particular illnesses,syndromes, or disorders (e.g., ulcers, premenstrual syndrome, dis-ability associated with tinnitus).

Linkages between self-rated health and marital functioning werereported in a number of cross-sectional studies. In two studies thatincluded women only, married women who described their rela-tionships with their partners as more rewarding reported fewermedical symptoms and rated their health as better than those whowere less satisfied (Barnett, Davidson, & Marshall, 1991; Thomas,1995). In addition to these cross-sectional data, one longitudinalstudy of married women showed that marital harmony was asso-ciated with better sleep and fewer physician visits (Prigerson,Maciejewski, & Rosenheck, 1999). In two studies that includedmen and women, higher marital satisfaction was associated withhigher self-rated health for both men and women (Ganong &Coleman, 1991; Ren, 1997). In contrast, among couples in long-term marriages, the relationship between marital distress andhealth was stronger for women than men (Levenson, Carstensen,& Gottman, 1993). In satisfied marriages, the health of wives andhusbands was equivalent (based on a health symptom checklist); indissatisfied marriages, husbands reported fewer mental and phys-ical health problems than their wives. Overall, these studies revealclear links between self-rated health and marital functioning formen and women, with one study revealing a gender disparity.

Among spousal pairs in which one partner had chronic tinnitus,patients' scores from the DAS Cohesion scale, along with theirinteraction with depressive symptoms, accounted for 15% of thevariance in disability ratings beyond gender, tinnitus characteris-tics, and depression (Sullivan, Katon, Russo, Dobie, & Sakai,1994). Patients whose high levels of depression were accompaniedby low marital cohesion reported the most tinnitus-related role

-s;

1

"4Jt

1os

S•gS^J

S3

£$

'£

^

WJbOP

1C

'crt>

JH<D

&>

<U

3

Dep

ende

nt m

eas

W3

ii

nd i

ndependent

mea

su

rtc:op

1

T3

§ s

iiV5 "O

1

c rt

cfl — 3cx=3

•a'iS sCL

•o3

^

1

§

asso

ciat

ed •

tom

s.

«j £

§ 0le

r m

arit

al

satis

fact

-em

enst

rual

syn

dro

i

.2? °-S

tial D

istr

ess

stio

nnaire

tr °

1°"

Mio

nal;

Marita

lct

ion In

vento

ry

1> 'I,U5 (XI

o5 <fl

U

||8*| ii*^ <9 ^

'§ JlOs c

"O

_

Os

«

J3jnob3oU

5

|m

c .aO W5•c >.J3 -c

^ .w'1

||

1 6

s|

is^ a,o « MIE!= s s,§ f s0 0. &

ts12H

al S

ympto

m C

hi

CJ

1

;tio

nal;

partner

role

• (d

iffere

nce

be

twee

n1

and

conce

rn s

core

s)

OJ Jj ^ — «3

^ 3 «

U

CO

SE|

" elS^ "S > ^

o 3 |O 3

o M

=%i

G _

S ONt3 '>•J? ~^

-•••S

1 00

g 1 its 1 | l-a 1 g | _ l | ^ l |

111 iJil ,1 Ii Jflsll ll'B i!'|lif IHIIlI U ll&iiftli? ill I IHiS^ s ^ ^ f e s ^ f 1 3 o-s i ^ - g S l s i " ' ! ^ $%-* s ^ s S g^ 3 g 5 3 | | ^ ^ S |s |o|||,^§|^| 0-35 ,0 -a« |^ i

| J| | | | | | | § | 1 I | | gi 1 | g | | | 1 | | § | s | I

||^|Sbl|l |1 !"lllll 1 i'll'llfl'S !.sf^f

S l s ^ | ^ l l > -sl § > l s w ^ | l l & | S ^ ' 2 § S l | |8§llll^l^l.ls Illallilli?!!?! IllMf c u ^ ' S M S § < - i o 6 « - i « _ g ^ y S > - S - a 5 > . a - ; 3 | l . S § ^ S 3 T 3

lillil8^'!!^ I|tf.i^-sli6|.i!i| r i l | f .J|||1|||1|J|1|| ||5||

u 2 « ^ S o S ^ n ( o a , ^ « * ^ c u x t - s . S - ^ b o o n . S — . o g o - o t ^ S t c x c x ^ o a 1

fTM fi^ | "] j™ ^U ^J QH

z 1" o gi-i-ag>x-

< - - o f a i = - S 8 '« '" ° §5"S^ o i g § i c g -S^ . - ° .a -api — ' - C j s ^ S ' S - ^ - S ^ v- T3li

£13a e H - a S S - S '* " 1 1 =^ -^ S usll ^^£ i ! -o ^ i -s lc f 1 §•£t|S J-821I3J? 1 1 'il § 11r^*^ C ( 3 f £ c - c . _ O T ) « Q, fi C in 3 ,J3

^:J li'lll! ! Ifill i .1*2|S *O wa t ^ ^ ^ S ^ i ^ c " ^ " ^ C l * ^ O 4* "O

X £ tx U S (S H

*o^ S >»

1 - 8 ^ Jll ii jj|ll1£ i« ^ S slll^ || Sf^g-rA =*• ' ? M ^ 3 ^ S ^ ^ - O ^™ g « 0 . a

11 ll. III if |ll Si. lilt1! Ill ||f I llL til Illl

QJ *^ u f?r\ y ^ ^ " W S o W ^ ^ " S ^" gj ^ ^^ S s / ) g 2 3 o j ^ J S ^ ^ H ' ^ - * 3.,^ '"''— T 3 U

<o c S'iSJi '5ic^ S ^ S « > c 'SB'S.™ S § " M

U O J U U J rj

O ij

•g O 0 ^ L> 'S -n C "* 2 C o c * - s j ^ — 0 ^ . 3 — H TJ u r S B

f^gl^," fl is l i^S || -a 11 elf i|R S — o "S II c S * " i S t » ° ' ^ ^ ' " 6 B S ' ^ " " S JJ S

iilfli ?!! Ill ii!t ^ c O O ) ™^ r t O i - O ^

a ^ c g g 3 - o 8 - c o g x o 0 l @ l ^|1 111p .§ 5^ 9 e S "51 ^ " S ^ p c ^ ^ ^ . H j i " c^ ^? P ® "

ON~'M f N t / > t r l » O O N ^ °1 ^~— (N CO — ' Tt \O '-H

s" -sON ^ <^ C/5

~ ^ S B- g *J

c g^ >» g S p-£ u! - O N S r < ) ^ T - } - JSD S\ ^ I/} C^ ON ^*"3 • ~ ^ O N ^ ' ~ ~ ' j 3 ^ O O N $ t/~iu JS ~ S | 0 £ ". * S

M Z_ g .U 0 | C -^ | ^ jg

i ^ / § ' O r § " n — <*l " rt ^ ^ QJ ^ qj J2O b S _] c« J H

~§3

.C

CO

(N

.O33H

c'•Bc

c<xl>£'u&

WsD(/!03U

G

11OJQ

«a

cd

sc^"2c/3

c03

1.22

I•O<L>C

C O

s^> •o ca.s Es .£>>> D-

« 3

s

T3 D

.c ^

Jill.~ "5 ^ "3 <u

g_"o 3 S '3

° 1 Q 1 .iu " *C '-a go •§ _Jj 5 A

^

1o

«

ll

11.S ™S "c? .2?

O

•S Al

ii 6 .i, as|!fc 'o .2? s

1 | £ ~

cs -E Q — —r~

\D

3

'&3

^

4J

1a:rE

a:

*-»«

ores

and

low

er c

onf

DS

sym

ptom

s.

" E

Q |S^§•85 a';> H

C «

II

*

O

&0 Cll"e >^

^ U

U I

8 "u •*>Q^

0.

•i

^"1

ll

Cro

ss-s

ectio

Inte

rper

soIn

vent

ory

i

5« ^^•§ig a§

1^8— ?U

r-

*~c'

1oOc/i

cL ^2 ®• — 

o =3 ^S g> S

^ ^ v^3 "" c

ll i-

qj

O. O

|||^ '> ^

f"S -û I, c S >_au o

DS

"o c

c c• : u

.. -oc .&

Cro

ss-s

ectio

rela

tions

h

—c _

~ s •- !

°l|£w ° X, °

•1^1^J i 5r-

r-oo\

gj"

«

,o>>

"8 -ilX 0) 3 <SM 2 0" a!

•- t S -a "«

!PP|a|||

Igflljo E -f § -S -3eo 22 i> t .

.a s J _ 8 J>

.E & | — g _cD, Q 5 " S f

° -2 5 'S " ^>

1 1 I | 1 1«

1 ô u o

rt t/i X3 *o3 4>U 3 00 t« w

!_, g j> J^

o 3 ^^ ^- °

_" -g .| §

'5 "« ^ £ "c4

S « S 2 >%

B

c^ <u^- >

« lo

III. « ca g o

^ H a1 U TJ.S P <u•o a E3 8 E|ii

ô

oiJiiniifio ;g u ^ c ta

§ "g | •« § -g

1C S If j- II 1m

l_'oBOCou

«fc

N

CJO

P— J

ILo

^

u«

JD U,ajT3 > w5

Sll11 IIl511 =-^ C (U

™ OJ °13 > iS•r 'S c

H 2- "S•S -§".

-S | c• — jJ rt "O

G r£ -5 *S

O "« O. en

s

1o

s^251&

jV*OJ00

Q

0

1

1

E

1 '-3s a gl|l

"S

(D C (UE 3 £$ w «^ i-S,•S^^

1 ' Sr- U £

12

11 ONS1 ON

c5 ••S "o

llSO rv*.

<X

S«0.S

AIM

S =

Arthr

itis

I

&ocuc

c'c3O.13|cuE3

1

"«

!cu>P3

W5

£ •> S

« 1SiS£

|i. r. GV 3jlj (£5

iffQ

dJ c

llJD -O

CII M

-LT cd

r H u-u '5c oc •&B u•i, n•o

S E•i^

<Û "3 y.

"5 -2

C '>(D 'S£ °t?,D 'I

^ <•o . -C3 W5

f^1 oII C/5II J

t/5 C

Q 1

s «fe S


disability. In addition, spouse-related punishing responses andtheir interaction with depression explained an additional 15% ofthe variance in disability beyond gender, tinnitus characteristics,and depression. Although both male and female patients wereincluded, gender differences were not assessed.

The validity of self-reported health data has been questionedbecause of correlations with psychological distress (Mechanic,1980); this argument assumes that distress leads to spurious over-reporting of symptoms, one important methodological consider-ation regarding studies in this section. Indeed, when respondentsrate diffuse symptoms such as fatigue, correlations with depressivesymptoms are often high, an unsurprising finding because of theoverlap in symptomatology. Nonetheless, global self-rated healthis a robust independent predictor of mortality (Idler & Benyamini,1997). In addition, self-report methods that focus on very specific,well-operationalized symptom clusters can show reliable associa-tions with physicians' diagnoses (Jenkins, Kraeger, Rose, & Hurst,1980; Orts et al., 1995).

Longitudinal designs enhance researchers' ability to control fordistress-related variance in self-reported health. One of the mostintensive assessments involved couples who provided data onmarital quality and illness symptoms annually (Wickrama, Lorenz,& Conger, 1997). Participants with higher initial levels of maritalquality reported fewer physical illness symptoms at study entry.Moreover, improvements in marital quality over the 4-year periodwere accompanied by decreases in self-reported physical illnesssymptoms.

Another study also highlighted the value of longitudinal assess-ment. Among adults in the Alameda county study, the associationsbetween peptic ulcer and marital strain were greater among menthan women in cross-sectional data (Levenstein, Kaplan, & Smith,1995); however, these findings contrasted with data from the 8- to9-year follow-up in which the prospective associations, with thegreater control afforded by the longitudinal design, were actuallystronger among women than men. Of importance, these longitu-dinal effects were significant after controlling for chronic diseasesand health risk behaviors that are key risk factors for ulcers(Levenstein, Ackerman, Kiecolt-Glaser, & Dubois, 1999).

With regard to gender differences, some studies revealed genderparity in marriage-health linkages (Fisher, Nakell, Terry, & Ran-som, 1992; Ganong & Coleman, 1991; Ren, 1997; Wickrama etal., 1997), whereas two suggested stronger links for women (Lev-enson et al., 1993; Levenstein et al., 1995). Although the smallnumber of studies that included both men and women and assessedsex differences hampers strong conclusions about gender from thisset of self-report studies, none provided stronger evidence for men.

Some of the same methodological issues that characterizedstudies of objective health status indicators remain problematic instudies of self-reported health (e.g., absence of normative datafrom marital adjustment scales), but the evidence nonethelessconsistently suggests that marital functioning is reliably related tohealth. Clearly apparent in cross-sectional studies, such associa-tions also held in longitudinal studies that provided greater meth-odological safeguards for the potential impact of psychologicaldistress on self-reported health. Overall, the studies in Table 2reveal that self-reported health is associated with marital qualityand cohesion, and also with spousal behaviors that are perceived asunsupportive and punishing.

Pain

Pain is a pervasive medical problem, accounting for substantiallevels of disability and contributing greatly to the overall burden ofillness (Turk & Melzack, 1992). Pain can be accompanied bynotable changes in physiological functioning. For example, dys-functional alterations in stress hormones and/or endocrine stressresponses have been linked to some chronic pain syndromes(Geiss, Varadi, Steinbach, Bauer, & Anton, 1997; Jones, Rollman,& Brooke, 1997; Lentjes, Griep, Boersma, Romijn, & de Kloet,1997). Moreover, pain can provoke increases in heart rate andblood pressure, enhance secretion of stress-related hormones in-cluding catecholamines and cortisol, and suppress a range ofimmunological activities (Kiecolt-Glaser, Page, Marucha, Mac-Callum, & Glaser, 1998; Liebeskind, 1991; Pezzone, Dohanics, &Rabin, 1994); indeed, acute pain is used as a stressor in bothanimal and human studies for this reason (Greisen et al., 1999;Page, McDonald, & Ben-Eliyahu, 1998; Santos et al., 1998). Notsurprisingly, anesthetic techniques that block transmission of no-ciceptive impulses can significantly reduce neuroendocrine or im-mune responses prompted by extreme physical stressors such assurgery (Kiecolt-Glaser, Page, et al., 1998), as well as mild acuteexperimental pain (Greisen et al., 1999).

The studies presented in Table 3 provide evidence that maritalfunctioning is associated with pain and pain-related disabilityassessed by self-report, performance on physically taxing tasks,and objectively coded pain behaviors (e.g., verbal, nonverbal, andfunctional expressions of pain experience). These pain indices maydirectly reflect nociception triggered by underlying pathophysio-logical processes as outlined above. Thus, the marital relationshipmay provide one direct route for maladaptive physiologicalchanges that contribute to pain outcomes. In addition, to varyingdegrees, pain outcome indicators may simultaneously reflect mo-tivational factors or social contingencies. For example, spousalbehaviors such as pain-related solicitousness can reinforce mal-adaptive pain behaviors, thereby promoting disability (Turk,Kerns, & Rosenberg, 1992). Compared with the objective andself-reported health outcome indicators reviewed in the prior sec-tions, pain outcome indicators reflect a unique interdependence ofpathophysiology and motivational effects, and thus we chose toreview these studies in this separate section.

According to social support models, encouragement and otherpositive, attentive marital interactions are typically viewed asfacilitating adaptation and enhancing health outcomes. In contrast,a number of pain studies based on an operant model have providedevidence that such spousal behaviors, particularly pain-relatedsolicitousness, contribute to maladaptive outcomes (Turk et al.,1992).

In a comprehensive study of patients with chronic back pain thatincluded relevant physiological variables and a group of controlcouples, Flor, Breitenstein, Birbaumer, and Furst (1995; see Table4) contrasted pain perceptions and attendant physiological re-sponses to cold pressor tests when the spouse was present orabsent. In accord with the operant paradigm, greater spouse solic-itousness was related to higher pain perceptions in the spouse-present condition for patients but not control participants, and theformer also reported fewer positive coping self-statements. Theheightened systolic blood pressure responses that characterizedpatients' responses to a cold pressor test while alone were not


exhibited when the test was repeated in the presence of moresolicitous spouses (Flor et al., 1995). However, increased bloodpressure, a hemostatic response, is associated with reductions inpain. Thus, the authors suggested that the absence of blood pres-sure increases in the spouse's presence indicated a maladaptivephysiological response.

In another study of patients with chronic back pain, higherspousal ratings of solicitousness were associated with less timespent walking on a treadmill in the partner's presence than whenalone (Lousberg, Schmidt, & Groenman, 1992). The perception ofpatients with rheumatoid arthritis of overall spouse responsiveness(including solicitous, distracting, and punishing responses to paindisplays) were significant predictors of pain intensity and painbehavior assessed during a standardized battery of physical tasks(Williamson, Robinson, & Melamed, 1997).

Certainly it seems reasonable to assume that marital adjustmentwould influence pain outcomes and spouses' responses to painbehaviors. Indeed, associations between pain outcomes and maritaladjustment may be moderated by gender, and these linkages maybe masked when analyses are conducted across gender in mixedpatient samples. For example, among patients with chronic low-back pain, marital dissatisfaction was associated with greater self-reported pain and disability among female patients, but not males(Saarijarvi, Rytokoski, & Karppi, 1990).

Alternatively, marital adjustment and spouse responsiveness topain behaviors may represent partially independent components ofthe marital relationship that bear unique associations with painoutcomes. Indeed, Turk and colleagues (Turk et al., 1992) offereda reformulated cognitive-behavioral model in which they sug-gested that marital satisfaction moderates linkages between spouseresponsiveness and pain outcomes by altering patients' appraisalsand interpretations of their spouses' responses to communicationsof pain. On the basis of the evidence that relationships betweenspouses' responses to pain and key aspects of the pain experienceare prominent only among satisfied couples, these authors arguedthat it is only in the context of global marital happiness thatsolicitous behaviors are perceived as such. In contrast, in thecontext of a dissatisfied marriage, patients may question the mo-tivation of their spouses' behaviors or appraise and interpret themas unhelpful. Indeed, their own and others' data support thisconceptualization. For example, in a sample of female patientswith chronic pain, higher marital quality was associated withperceptions of the spouse as more solicitous and distracting, andless punishing, in response to pain behaviors (Kerns, Haythornth-waite, South wick, & Giller, 1990).

In addition to the role of spouse solicitousness-responsivenessin increased functional impairment and higher self-reported pain,the management of marital disagreement may play a role in painoutcomes. For example, premature termination of a physicallydemanding bicycling task was more likely among male patientswith chronic back pain who first discussed a conflictual maritaltopic with their spouses, as opposed to those who discussed aneutral topic (Schwartz, Slater, & Birchler, 1994; see Table 4). Inaddition, marital conflict between male patients with chronic backpain and spouses was associated with subsequent increases in painbehaviors, which, in turn, appeared to promote more punitiveresponses by the spouse (Schwartz, Slater, & Birchler, 1996). Inlongitudinal analyses of a predominantly female sample of patientswith rheumatoid arthritis, negative spouse behaviors such as crit-

ical remarks predicted greater future pain even after controlling forbaseline pain (Waltz, Kriegel, & Bosch, 1998).

Finally, three studies have examined marriage and pain out-comes in the context of psychological interventions. One series ofstudies examined the impact of participation in a couple's therapygroup in which five sessions focused on communication (Saari-jarvi, Rytokoski, & Alanen, 1991; Saarijarvi et al., 1990). Amongthe 63 patients with chronic low-back pain who took part (35couples refused participation), there were no significant changes inpain, disability, or clinical assessments. Of importance, however,there were also no significant changes in marital satisfaction, aperhaps unsurprising finding when the intervention was relativelybrief and couples had not been selected on the basis of maritaldistress.

In contrast, results of a more intensive spouse-assisted copingskills intervention administered to a group of male and femalepatients with osteoarthritic knee pain showed that patients whoreported increased marital adjustment from pre- to posttreatmentwere characterized by less physical disability and less pain behav-ior on completion of treatment (Keefe et al., 1996). In data fromthe 12-month follow-up of this intervention, patients in the spouse-assisted coping skills training (CST) condition who reported en-hanced marital adjustment from pre- to posttreatment also hadlower levels of physical disability and pain behaviors at follow-up(Keefe et al., 1999). However, higher marital adjustment wasassociated with increased pain and poorer coping among patientsassigned either to CST without spouse involvement or to anarthritis education-spousal support control condition. The authorsspeculated that individuals receiving CST might have initiatedmore candid discussions with their spouses about pain and coping,whereas spouses who were not included in the training mightsimply have increased solicitous attention to pain symptoms, in-advertently reinforcing pain behaviors. In addition, participants onaverage were not maritally distressed prior to treatment, thuslimiting their range for improvement in marital adjustment.

In summary, the studies reviewed in this section provoke novelquestions about the role of marital functioning in physical healthoutcomes generally, and in pain, a costly and prevalent healthproblem, specifically. For example, spousal behaviors that appearsupportive on the surface (and that therefore might be assumed tolead to positive health outcomes) may actually culminate in com-promised functional status and more physical disability. In addi-tion, pain outcome studies provide conceptual clues as to howresearchers might meaningfully model the multiple facets of mar-ital functioning. For example, global marital satisfaction may serveas an interpretive backdrop, altering patients' appraisals of specificspousal behaviors and, thereby, their functional significance fordisability (Turk et al., 1992). These data suggest that if models ofmarital functioning and pain, or perhaps other health outcomes,exclude either global marital quality or specific spousal behaviors,their explanatory power will be compromised. Beyond these novelcontributions, many of the pain outcome studies reviewed herereveal methodological limitations described in earlier sections,namely, (a) the inclusion of only one gender or lack of attention togender differences and (b) the paucity of dissatisfied couples.

Although there are not compelling gender differences in theliterature on marriage and pain to date with few studies makingexplicit comparisons, there are consistent and notable gender as-sociations in both acute and chronic pain; these pervasive patterns

>•

1ô•S3a.

8-QS

2"s'CQ

'J-.COc•5c

«<uCJa:

D

S£

1ex

c

1a11llc50

S

_

"iIT3

fl C

||

§.oCCS0.

>>TD

CO

a i

3ii w

as a

ssoc

iate

d w

ith g

rea

d re

sulta

nt d

isab

ility

in

nen.

Mar

ital

diss

atis

fact

ion

'

tal

diss

atis

facl

!f-ra

ted

pain

a>m

en, b

ut n

otC3 ^ *

S

"a J*

o 'S S"

^ " JT

-M '^ O

|||

|||

*

„

2 D 1S 2 -

•3" S

111" i s|5.lu

^ g,* S3 1S « c'•5 j^r si°" 5

s 5, Ss'i..|G -^ «

" ^ —

" s 'E£ _o g~

1

_, ~,1 I:?^

;al

exam

inat

ion

findi

ngs.

ês

in p

ain,

dis

abili

ty, o

rits

.

~ o ô s 8

— '£ ySi .SP'S3 c u

1•3 J1

'S. 'S ^J ? "". ^ T3 ™ ^^* £ "s

l|Il

^

^1 JE g-

— a

ongi

tudi

nal,

expe

rimen

t;se

ssio

ns i

n

j

S &* u |S OT O

' fC" £^Q, Tt g

c ^r c« ^ p

'||C ^ «^ C5 ^3

" S 'i£ _o g

—

0

o ™~

•— ' c'

f^

.s

/hos

e sp

ouse

s w

ere

mor

ei

mor

e pa

in a

nd w

alke

d le

s:

-£ c

•£ S's |

'S ^(X

o "a "

a 1 1'5. n g

o3 fe ^

•§ S 'C>- •§ ^

"f- |-g

1 af

S

•So. §•> D ^

§•1 S

jni} " O ft,^•e G "a l ¥ §D C t« o

U

(N

^ =D- ^ "^~U ||

0 "

1 1x: c

• o

a5 T3S ^"

1|i>i

•5 ^12-C

^ i<B K^ o

|°J

¥

cKJ

^%

1VI

i pr

esen

ce o

f

•S

|<u1

2

men

t d;

I«—i

w1

I1 =3± SS °-3 <U

§-§en E

liciti

ous

spou

slic

itous

ness

wpa

tient

s.

•J: V. O

ll

~ g |

trea

dmill

te

WH

YM

PI,

by b

oth

pat

<5 -n °•S S «Sf 1 s-Sl s.a & a -S 1 s 1 - -s .s *8 ^ ' S s l . l l a S g j . S

^ c - ^ l E . S o e ^ - s - SS ' 3 o S s — ' a - ^ M C ca °- c 'S ^- "2 '0 B " .S '3 «1^.2 &s ,a=§ a- js *iS ^ e « J u « ws ^ T5, -2 o. S r-E I -111 1 y|'§ a ||c "7? £ £ •— "-1 ooo o.S >ic

1 § 1^111^11

ffliiifPiii^ > ^ "u j: cj 'S • — • "5 & f-. <Aa. £

- & 1 ^J | | c | |

2 -g Q Q. ra t>

J £ £• ^ c'J,

|l i l||1 1 M „ 111

.£ ™ i- co -^ x> ot2 m

•S1 u

ll gllil'l^-"11 -.- s ^ lg I' S ^ W ^ ; 'S «J C "O K ft<

a i O a j C S o . . < L > u - S :

I! i!£ IIHI10 .^0 .1^^ ^ n ^ - c o oU U

c c —D Oi 2e gj g |

^ > C ^ oC ifl Q C C

& Co1 B °- f-^.y ^ X •— ^*c i i 2 ^ i is " ; | "

1 1| 1 1|(i ^ a c ^ u

§ — u w~i E

il| iS|co rt C O rt ea

SON

^ 2 1«>• tf 5C ^

•g« |

* c

spon

ses

wer

e as

soci

ated

wit

vior

s am

ong

patie

nts

repo

rti

;sen

t.se

sol

icito

us r

iate

r pa

in b

eh

Q- O OX)

"

«J

ul^ J3

* 'o

I'|

" "§

I ito

•lat-£ro

ss-s

ectio

niB

ehav

ior

C

U

•s "S* 3c I

3 "§•0u ^t

1 II

il« a.— .y

^ -ioin

£

—

ao

s_

ity. S

pous

e so

licito

us r

espo

rha

vior

s w

ere

asso

ciat

ed w

it!am

ong

patie

nts

repo

rting

i. R

elat

ions

hips

bet

wee

n

»her

pai

n se

vepa

tient

pai

n b

:ate

r di

sabi

lit)

:ate

r de

pres

sic

js S oo w

.2

Jl~ u3 "

11

"»."%

11

O

~ £ £

|S|<2 -s a

H

,

CO<

ioD-

1

M .fa

resp

onse

s an

d pa

in b

ehav

ioi

mor

e sa

tisfie

d co

uple

s,i

patie

nts

who

des

crib

ed t

ht%

in h

igh

leve

ls o

f ne

gativ

e

licito

us s

pous

t;re

stro

nger

fo

tally

dis

satis

five

s as

eng

agil

t/> > c3 >

"

4J

U

I'iS "SCQ "

155f|

1 6

5

—'O

U

^ S"1.1O .S,

o ^^ "«I'i

1 §a*"" C3

. .

"« U"l

S II§

1cr

^

1'S

*S >,-m

spon

ses

to p

ain

and

low

lev

repo

rted

grea

ter

pain

sev

erit

•s.

Am

ong

mar

itally

sat

isfie

i; b

ehav

iors

wer

e as

soci

ated

oo <5 > «f I J j s

W O M D-

•85&

1

1(Ull11nj 5C00'^

D

ge)

stm

e

g sE S2

^y1 3 * . & =

•S a. feb^ £ -S ££ S,

•§ Sal S.&1§ o ' j . g ' S e : S

iil lei 1• S - s a a l J S .al-s l

a? .1 '1 so a

!"*=E ^

P1I ra

O cu ^

S S . S

•

° gII ,»ll

,- 1 §§* B - - E

--

11

a -a-11;-^ «.

II

Js ii

i.

of p

ain

inte

nsit

y or

unp

leas

antn

ess

betw

een

cond

ition

s. M

AT

sqo

res

wer

e no

t co

rrel

ated

wit

h pa

tient

s' or

spo

uses

' ra

ting

of

spou

ses

solic

itous

, di

stra

ctin

g, a

nd p

unis

hing

res

pon

to p

ain

beha

vior

s. P

atie

nts'

ratin

gs o

f sp

ous

resp

onsi

vene

ss w

ere

asso

ciat

ed w

ith

grea

ter

dise

ase

impa

ct,

pain

sev

erity

, an

d in

terf

eren

spou

ses'

rati

ngs

wer

e no

t as

soci

ated

. N

o ge

diff

eren

ces,

ighe

r an

xiet

y an

d lo

wer

mar

ital

qua

lity

acco

unte

d fo

r 21

% o

f th

e va

rian

ce i

n w

ome

X

c/°'gS

S |

'i .a1 s "a .-E "> S £, «v c c J2 c

o • - i SIIf! s d

^>

il|J|i Jl l l l l f l I

(J

'£00 3(N JJ

K 1>k

«5

_^ CJ ^

lSc IfiS- 1s-*" -O

(—

£'

"3

| ^.i

1 £

1 |*

pain

rat

ings

,sl

ativ

ely

cons

iste

nt c

orre

latio

ns b

etw

een

paii

o/

OJ

1(X

•£*«

<U

X£rt.cCJJC

O

<U

3S

"«3

•53

'5bc

i-J

—•£C

T3

heum

atoi

*-.c'ic.u

IV)SO

OCO

•3

1)

•J

1

U

mea

sure

s an

d ne

gati

ve (

but

not

posi

tive

) sp

i

c

1

1CO

IG.

•g

e

12-o

SB

H

o

_b'"3

0

r-~

I I

oo

1

beha

vior

. In

lon

gitu

dina

l an

alys

es, n

egat

ive

c'I

Ou

a?<U

RJU

CO

and

nega

tive

posi

tive

aT3•£

I

3

¥"2?3

stan

dard

ized

:sp

ouse

beh

avio

r pr

edic

ted

wor

se p

ain

outc

oev

en a

fter

con

trol

ling

for

bas

elin

e pa

in.

No

« "§•o u

* c2?!,1 -ac So -jz•3 «3 â £

nter

acti

ons

Q.tc

j>,

10)M

'o

itien

ts i

n sp

ouse

-ass

iste

d C

ST w

ith

pre

- to

post

trea

tmen

t D

AS

incr

ease

s ha

d lo

wer

lev

t

£

c'>•§ I?

1|

f j11CO S

Q

O

ftts s

•S .—o a

=3 °-S.

•Q% £^ 1

| Isu ^

1 "| §

mal

e an

d 54

oste

oart

hriti

s

(*-i

g

—«4;

<u<D

n

phys

ical

dis

abil

ity

and

pain

beh

avio

r at

fol

li

£

"i1

HCOU"O

1u

a.

IIW5

Q

CO

up.

Pati

ents

in

conv

enti

onal

CST

and

AR

E-

th n

o sp

ouse

CST

wii

wit

h D

AS

incr

ease

s sh

owed

gre

ater

pai

n.ne

nt,

or A

RE

-ro

l co

nditi

on

« c^ oo yc «>.£ to

S2

3

£

I I

>

2c

—c£13ccd

c

3 _QJ

_4> Oâ „

=7 SOJ C

i >s-3

s= =I I E__^ ra

CL

K 3

^-T "C• ou^ G-

1 A

djus

tmen

t "

>n-s

pous

al s

uj

llsÎI w

f- S< bS c

c X3u 03B °J

S <

3 S1

'5 !§|S

liM tÔ

'C c

IfI I I IU H•§ tsi,u

1 sS caj O

~ c.2,111•o .Q 'C

"« U.Q.S" C

Q Î I

5: J

: _

^t

1

aQ"a'5s

"S

I=X,-c•t2

^3S

s;• 2og

-S;*»i'gC

MC

1t£

e<tj>_ui>

C*

Dep

ende

ntm

easu

res

<u3

1

cu18.1|C

1

^ __ u

ll3. 3

c <"

S.2'§ 'S1 1o.

~a3

V)

i

!co1ajc

o£

u

Q.

1

O

~5m0

.5 Iu jy

-C i£

C^ ra

0 g

g •=

1 -o '^

0 c

1 1fl~z

aI^ a,'s m

o «

*

C

£

O

iTc —"1 •=

llCL, W

*-• 1§

|*¥ f.. >

i-sect

iona

nfli

ct,

rec<

o 8u

r~

ilu -^"H.go '° tr-i

OJ

^

"o.D-

c^J

"S °O "~

r~t~--' C

u «j

= +

6.6

BP

M;

SB

P i

ncre

ase

8. 'S.&* ^S 6

i^lft i?

a

CO

00c•c3•a

S

10 ;_;

^ I•1 %8 _„.

S 1

1 1H g

U

&

•g^S« =

• S ^ c :« ^5 Eo c oi« C

CN

^•5,

H

IUJ

Is "S C 2 B S |S ™ 8 o -S « g »•§ 11 g o^ g g g 8 | = S ^ . S g B 2 . | g

Illl™ §lsli--3|i1I| Is -s £§11^1 |.l

s •= ? » 1 ll*|la||Kag.2.1 1 s^-^^ ls* 1 1 : !E § ^ . S ^ | §ll§ M± 1 | B

^ E '1 1 ^.1 B -o "§ ~ •! e ] §5ÎI fs .5? Q « -~ c ^ ^ ^ u ^S 2 a o '> i •§> "| j| ii > a -| .3 J|l|lf | s §g ||>|||

jff ljil|isjlH|;c « c | 0 5 u

^ _ S 0.

<

£r

^ > ^

O JZ "*~

1 J5 a s |

! In!!'> C X! *O o c

1 y ||Sl

~ .. o §

1 l! 1!| CO & | « |

6 j_- "? s § ™

^ u" ? •= i 'i. °c «jj o s i i

u

•o ^>

I^|s7|^s^ S «« I' "§ * <N II

B^ ' ^ u^"* II H

a1^ '" 2 ^ S i ",, , « ' 5 3 c ^ sg

_j \^

<* S S3uT 2 ^

£ af O

II ll2

•M i!II .s ^fl §•! 1

Hillfill!

Hoofs j?^^ *5 .— ^ o

.S.H E S.1 1

fl? °1 suiiiiii

u

1 c S t- O O «>4J « -5 >.—s. c S .0 â

CQ

c_o

ll"a S1 =Sc *2

S C

y 1S S03 u

lli.?^

1 1-538

O.'S In 'S S

§"" U " WVi "° <!

^" *o ^- « Qs

.,

' "*IS55 trN |2

|I

1)

•a .2C 60 H™ C «

l||i

w -g c 3

1 §>! 1

fill

>> S "I ^"™ "3 c 3 a. « _«'55 i- -S 3

"§, C O «

If If.£ £ S U

c u

^°- glS < 1 S M

fclll—

32

si

ll

U "f

?! .s||CO "?

.. 0

I'S0 "

i -6u

u S II

flfljleg C .H ^ 1 JJ CO

"&'C 4J pj £ > H

Q Q, H -O *- C

° (5 "S M *3 C C

S

c£

—_•s£

1

SS U

° 0) -0 § 00

111! I0 _g .S

s * a .2

t|| |I

• O y c S . | a. .2 S 'I "S

fl.18 ^11^^ S " 1 - j - o S l H c 0 2

i 'i c E §o ^ 'u > E'> u .§ J3 <9

ll^'* 1

fl: 1^1 1z £

?!i^îJsii^.3 S

if 3= b. >

A *O *3

U .C '§,

c3 S

i^co ^.. JJ y J2

|J| J

tfl S o y

U UJ

S3 .,"E.P3 fO

S O r~- ^S

* N" II -

ê II H 1S w, •<! 'S

J c: < ex

$~ t

— c

§ i | 1it S. ~ §>> S3 S3 S-g S3 3 «

|o» g"S E

S J c i" °2^.2 u "«

I -c I X S

ill 1 |

S ? is j- 'o§ I 51=3u u c S %

C "u en -i

.2Sp1 s » i -^| S 13 S u

a

||1•— c« . -

.= B &-

2||

&• M E>> . 1>

is 11|| o"" 0 g

"o 2 " ooS 3 3 5o c & .5

S r 3

ll li« JJ It § §

"S 1 !5 ° w

§"" ,« u ~" "a.f*-, ^5 -a 3

"o C- £ rt 8

tr-"

1 2J ~lu HS £

S .£

- 1

for

patie

nts

bi

Jso

report

ed

f<

fi - 1£3 12 22

1. 1 »

1 P

ill

ill

11ts.H,"w

.cc1,

Tt

u3

SS

i'•3ci£c>

—u02

g S•g 3

8.8

1w:

UE

1

11-a

c.5?

Q

*o§ E2. §

a|e s1 If

fi

>.

1<55

cl "B o . s « fc 1 - 5 - 1

U ill! J tSS !1M§•1 S = I s i-*l -i S^ ^°ob » a 'o > a, -o * « 8 S u§8 |E2l ISIS s^lII lf:f l|ii l!ji11 ^||f 11 il Hflis Pl^ lll^'ills§1 ^l-l i-slllf!!1!g^ | s " * 8gJ5 |g | . § -ga£ » > *>,a: • " S u 5 1 - 2 ^ 0 ^ ""8.1 8-f£^ S - S - S j J ^ S gfe S e -g a 1 |S -gg | | l sSis - ^ K l 6 S i*ss l -§ iS|u s | fe^2 ^ l 2 - a ' 5 a « 8 oO B f ^ S u + • -S ° .« 2 T3 2 u -

Si a ^ ^ n l l - S g ^ S l a g l'« "S • ° a t . 0 _ y 2 § 3 § ' 5 a g f a i 3fls|llsiisisl|illZ X 5 H

x"ft U " £> 1

v-T 2 H u3 a. w ^ «* j* SE ca Z C i " I S1 cl Z S ° 1 = £

S i tu a.

1 1 § J = 2 - a . G S ..II <1|-S l|sl |-|ISw « r- u a" ^'S o i? 2 rt ^

™ S. " > ^ o d o ^ r a - a " 4

11 ?|*f J^ii M i ! §?1 glls |l »l1li0 8 G - o y S o s - s ^ s . y o _ . o - |S E a,"2 1 'S S l o E 2"S S ^ S f j Sl j ~ u i c C f l c 3 ^ & o S o _ ; C i . ^ . - i f f < L ) 5 J ' n o o

^ - 1 ™J 1 1 ^ 2. « 2 1 1* ^ !! !. , u *L S U w i " O ' p « l E ^ " c 3 ^ 2 -" - U

jlll||||!||.|-|l|||f'H^IH^H S 1 1 sl'liii 11J U U U

r— oo00 _' iS upi " * " §-s i $ m

J c * ^ ^ ^ '"'^rj f N " £ ^ < CQ S- Q *O - ^™ " 3 'j S C Q

C- OJO ^ sO C S- W

y, .Sb '« i f >," c 3 2 = o j ? S 3— S e "S.S * ii f- - g s S s|llli= bl Hill0 o o **-in Tt a\ *t

cv"2 "g "3 -r£ 3 S g 2

X « IB w>j J= 9 2 .E

^Hi "cS 87°^ EO•3 s .a ^ -g^ £ ^ en

p g •aG i." J .a

Bflll „PJ o 4, 3 >^

Z -a g g S- S -c E ss

£ « - - « •u-l

f!

<x

S^,8 U"

*5?

tj «5 C

8 SI

0 'C II to1 isgJ 1 " ir§"1-18 M^

OSX

E s

E =•S S'S 2

I

« 00

I

J= O

^-52? 8.S- u S

5 "i tl= >

'g C

IffO M- o

-Js.2

1w

1 '

^jyC3H

obc•3

c

JJ

Ctf

c 5S.3Q

11OJ

cOJ•o

1•3- —•gac&0

I

•oc —TO C_ o

a|-.i,c w

-- 1

d-

"o"3

^5

•o

pons

es i

n w

om

en t

han

lues

30

min

aft

er

confli

cl

in c

ou

ple

s w

ith

po

sitive

trie

couple

s and

dec

reas

e.

BP

did

not

chan

ge i

n

iflic

t.

•f. ^n «- O i- t-

£ > a E ^s gHills•e 5 g « -S a8 -S STo s; §

onfii

ct

pro

duce

d h

igher

men

, an

d w

om

en

had

iC

ort

iso

l in

crease

d d

uri

behavi

or

and i

n w

ive

sin

couple

s w

ith n

egatr

men

or

wom

en

in

rcs

p

U

1

gX?

Q.00

O o'C c.

s > i ££ E S .S- « S E&- rt £ ,1, ^-,£ u, ^ ^L'•5 o g ~ ?jo u * ii S• > • - ^= cd- • s ,« ^ —

il*l|Sglll& .5 > JJ <*>

sMif• - — ^ c 3— •—• 3 C ~~g E J y g

Hi1!£ 13 j= > c| 0..2S3u

oc

II u^ CL '3 ^<

S. fj S -1 10.3 a ts so ~3 i t; ^ ^ ^tj £vj Q, II £•

:So

*<*

" S S I1? '«

its ^ ^

j§ -= J:c; £ aJoi£

, anger

dis

pla

yed d

uring

reale

r e

leva

tion

s in

BP

,and n

um

bers

.

2- oo >,— .c •—

o * S-c ^ 3•= « S8 .2 g>

•s !xtjlge ".s,§if i g•<

s^ ^ j£

.ii p. 0 g i

^ -g _ o O D > i C

r O O U «3Q_

X

^ • •ir StS uo cH -,x« &>, -^Z u«•> g"Si" SS s i^ ^ o0 3 C,9 2 c

U aj O

£ -o =o c 'E— cd C» -g <A

3 —w J= ^03 ~ 8— * == - 1.2 ^ 5 > "-"OJ _ J=

III0 0 a

U

_^

' n)

" -1

— § _ g

y i> '§ Eaj "p § t^

*

&^I"I»u,' *B S\

—^ i -*•>. >,^

^ S u.O

was

ass

oci

ate

d w

ith

/es,

acc

ountin

g f

or

9%

onot

rela

ted t

o t

heir

•o 'i ilji|

igher

cyn

ica

l host

ility

ii

gre

ate

r S

BP

re

act

ivity

the

va

ria

nce

. W

ive

s'

h

X

a.CO

^ "c

i *H•jn

*O

53

^ o -g

0 1 ^U y . ^pi U > tj

s Ilio '5 w o^ S ° «

5 S 3 '-C

0 & | ICQ i« ^

•3 ul1 gs

|Iau

o"T ,s~u "

1 1"f

c'

Sm

,c '5 —to

a o;* - o . s « "

!?M « ih HISill! lirit iiliIII! IIP! ill!3 § °»T3 S -1 SS S "S 8 oTs'o= "S | :f "S S §:> "i 3 g &S 2

Hll !|s?i!ji|jlfi| Illl

3 ^ S M H - E ' f - S - S ' c ^ ' - i ^ a : ' S ' H . o ' 5 2 ' - J s S | " S ) ' 5 2 ] - ! = ^1 111! iihil llli lllSt^li ill!1 | - l ^> fe:i|S'|?glM.s|-ot1§p 7e-.il! |!«! igf

CJ O U l£ « u >3 "O (I .-

•- 3 -. is 1^--II 1'!^ a. o1--ii h ^ §o- o. Si E c

CQ 03 S c g .2

"c "c CQ a 03 || § "

0- Oi Oi Oi" ^ I I K X g g. II c

8 -53 § a

^ u M "o S •—•' **32 '"" ¥ 'uo « JJ

^ g o | w ^ 1 § *§ 8 »

.s 1 i c & 1 ^ a £• 1 f J |_ f J a IT 7 i w£ S ~ a - 2 > 8 ' a - 2 > « ~ " i g s ^ ^ — cu — "

2^ § J l f e | J 1 S I "§ c" "i"^^'! S ^ 8 - |

]"5 Jllf |il|| ||| ||i |flt

H .H ra IM c c , — , ca t- c c * * y i 2 i ^ f l j c S ex S i £a 1 llS||l|i|| P|| HI! ||d|

- u « 3 W . j l 4 ^ 2 3 & ' . j ( l ~ l ^ JJ «9 ? ^ S -^ "^ " * ^ £

.2-^ t * " - - y - 2 ' £ l * " - - " S ^ " G ( * ~ c ^ c / 5 , l * - ' «u5^ . ^ - s c Q - a ^ j

!3 ^ P 3 u i « o j r ; ' " C Q u o c 3 j J 1 ^ ' C Q ^ ^ . S ^2 . - -1- "2 ' S Q 1 — £ ' ~

II ii*i!ii*iJii'!!| HI !| |j-i"l^ " r a S a . o ? E £ a . o S E £ - O D < ; S > - 4 c iS^^-^ oiji '^.t:

x x x x a - 3 c > cW uJ W UJ 7^ :« g c .0

§ a.^3 §1H1 ~o ^ '5 °• ^ s £ <.S -^ _u c ^

As AC; • c '"% >- M /^

C5 O « « II 2 S -— y?f t II OS II e3 c - - a > - J 3'u- , i » n T3 "a - S ^ C Q

fN™ fSH^ ^ _ * - > •§ — *^ ^ II X II

-|H fij g.|| g.|| |Q>3<

S| 8 S 8 g f 8 g f £ = " £ : . §un ui O O — ' ^ 2 >

o 5 S -cu C ^* ^" P3

«§ K ffi s -H

^ S u S «

<* S fl I g "Sco S §^ ^ m o -2 •-

.•i f- - » 1 o f 6 * 1 -aE " W ~ - T ^ * ~ ^ ^ ^ ^

• - - C J s W KI >— -rj # CQ -^

E'S "c3^j( O t / D ^ O la%J M rt O , d^ ! S ^ o l _ > g= Ov C C? J5 S) S •*= C " 'S %

£ ~ £ f f l 'i '§ o ' o X o ^ BCQ CQ W C/1 < O O C J t / }


strongly suggest that when marital relationships influence painoutcomes, the effects are likely to be more consequential for thehealth of wives than husbands. For example, compared withwomen, men have higher pain thresholds and tolerance for acuteexperimental pain, men are less likely to experience a spectrum ofrecurrent pains, and men report less severe and less frequent pain(LeResche, 1995; Unruh, 1996). Women may also be at greaterrisk for pain-related disability than men (Unruh, 1996). Further-more, as discussed earlier, these gender differences in pain per-ception and tolerance have to be placed in the context of themultiple adverse physiological changes that pain can provoke(Kiecolt-Glaser, Page, et al., 1998). Accordingly, any contribu-tions that the marital relationship makes to pain outcomes couldhave a greater impact on the overall burden of illness among wivesthan husbands.

Pathways From Marital Functioning to Health Outcomes

There is ample evidence that intimate relationships can impactillness processes or outcomes indirectly through alterations inmood, as well as through their influence on health habits (Burman& Margolin, 1992; Kiecolt-Glaser & Glaser, 1988). In addition,marital interaction studies that include physiological assessmentsprovide solid, mechanistic evidence of how marital functioningcan have direct consequences for cardiovascular, endocrine, andimmune function. We next consider evidence for these pathways,including the role of trait hostility in biological responses tomarital interaction.

The two groups of interaction studies in Table 4 assessedphysiological changes during and/or following a marital interac-tion. The first set provided data on physiological changes associ-ated with "conflict" interactions—discussion of couples' relation-ship problems. The specific topics were typically chosen by theexperimenter, based on each spouse's ratings of common relation-ship problems. In the second set of studies, couples were asked todiscuss and resolve impersonal topics on which they had beenassigned opposing positions, for example, layoffs in a hypotheticalschool district (Brown & Smith, 1992).

Marital Problem Discussions

When we reviewed findings from couples' discussions of rela-tionship issues, several obvious themes emerged. First, conflictclearly alters physiological functioning; data collected during theproblem-solving discussions differed reliably from those gatheredduring noninteractive baselines, and/or nonconflictual discussions.Most studies included cardiovascular data, and marital disagree-ment was reliably associated with heightened blood pressure andheart rates (Broadwell & Light, 1999; Ewart, Taylor, Kraemer, &Agras, 1991; Flor et al., 1995; Prankish & Linden, 1996; Kiecolt-Glaser et al., 1993; Mayne, O'Leary, McCrady, Contrada, &'Labouvie, 1997; Morell & Apple, 1990; Schwartz et al., 1994;Thomsen & Gilbert, 1998), with one exception (Fehm-Wolfsdorf,Groth, Kaiser, & Hahlweg, 1999). Patients with chronic back painshowed elevated lumbar muscular reactivity during a conflictinteraction, compared with a neutral discussion (Flor et al., 1995).Conflict was also associated with alterations in both endocrine andimmune function (Kiecolt-Glaser et al., 1993, 1997; Kiecolt-

Glaser, Newton, et al., 1996; Malarkey, Kiecolt-Glaser, Pearl, &Glaser, 1994; Mayne et al., 1997).

Second, negative or hostile behavior during conflict markedlyenhanced physiological change. For example, a marital problem-solving task produced clinically significant increases in bloodpressure among patients with hypertension that were specificallyassociated with hostile marital interactions; neither supportive norneutral behavior was significantly associated with change (Ewartet al., 1991). Indeed, the effects were sizable: Hostile interactionsand marital dissatisfaction accounted for 50% of the variance inwomen's systolic blood pressure. The authors summarized thefindings in their subtitle: "Not Being Nasty Matters More ThanBeing Nice" (p. 155).

Paralleling the link between hostile behaviors and elevatedblood pressure reported by Ewart et al. (1991), analyses of immuneand blood pressure data from newlywed couples suggested thatphysiological changes were significantly related to hostile behav-ior only, not to avoidant, positive, or problem-solving behaviors(Kiecolt-Glaser et al., 1993). Similarly, endocrine data from thesenewlyweds demonstrated the significant impact of hostile behav-iors during marital conflict on changes in serum levels of epineph-rine, norepinephrine, ACTH, growth hormone, and prolactin (Ma-larkey et al., 1994). Moreover, differences in the pattern of changewere apparent between behaviorally defined groups, with morenegative or hostile couples showing more persistent elevations onepinephrine, norepinephrine, ACTH, and growth hormone duringthe conflict discussion and for 15 min after the end of the discus-sion. Consistent with these data from newlyweds, endocrine dataalso evidenced significant relationships to negative behavior dur-ing marital conflict among older couples in long-term marriages(Kiecolt-Glaser et al., 1997). Both men and women who demon-strated a pattern of relatively poorer immunological responsesdisplayed more negative behavior during conflict; they also char-acterized their usual marital disagreements as more negative thandid individuals who showed better immune responses acrossassays.

A third key theme among the interaction studies is the relativelygreater physiological change shown in women; gender disparitieswere most obvious in relation to negative behavior. In data from asample of patients with hypertension, wives showed larger bloodpressure increases than husbands during marital conflict, andwomen's blood pressure changes were specifically related to bothhostile behaviors and marital quality, with these two variablesaccounting for 50% of the variance; in contrast, only speech ratepredicted men's blood pressure increases (Ewart et al., 1991).

Similarly, behaviorally coded negative affect accounted for 20%of the variance in women's systolic blood pressure during a maritalconflict discussion and 53% of the variance in self-reported maritaldistress (Morell & Apple, 1990); although cardiovascular datawere not obtained from husbands, behaviorally coded negativeaffect and marital distress were unrelated for men. Differences incardiovascular arousal during conflict discriminated wives (but nothusbands) in physically violent marriages from those in distressedbut nonviolent marriages (Jacobson et al., 1994).

Among distressed couples who participated in a marital inter-action study in exchange for marital therapy, wives responded tothe task with greater increases in depression, hostility, and systolicblood pressure than husbands; in addition, women's lymphocyteproliferative responses to phytohemagglutinin (PHA) decreased


following conflict, whereas those of the men increased (Mayne etal., 1997). Following conflict, decreases in PHA were significantlycorrelated with increases in self-reported hostility (Mayne et al.,1997).

In further data from the newlywed couples described earlier,blood samples acquired hourly from 8 a.m. through 10 p.m. werepooled to provide composite daytime values for six hormones;these data provided a way to examine relationships between con-flict behaviors observed during the morning's discussion and moreenduring or persistent endocrine changes (Kiecolt-Glaser, Newton,et al., 1996). Consistent with the gender differences in immunefunction, there were stronger and more reproducible links betweenbehavior and endocrine function among women than men. Forwives, higher probabilities of husbands' withdrawal in response towives' negative behavior were associated with higher norepineph-rine and cortisol levels; this demand-withdraw sequence has beenassociated with greater marital distress in a number of maritalstudies (Christensen, 1987; Heavey, Layne, & Christensen, 1993).In addition, wives who showed higher frequencies of positivebehaviors during conflict had lower epinephrine levels. Moreover,the magnitude of the relationships was noteworthy: Amongwomen, behavior accounted for 24% of the variance in epinephrineand cortisol, 29% of the variance in norepinephrine, and 37% ofthe variance in prolactin (Kiecolt-Glaser, Newton, et al., 1996). Incontrast, none of the six hormones were significantly associatedwith husbands' behavioral data.

The generalizability of physiological changes observed withthese younger couples was assessed using a similar laboratoryparadigm to study endocrinologic and immunological responses tomarital conflict in older couples who had been married an averageof 42 years (Kiecolt-Glaser et al., 1997). Among wives, escalationof negative behavior during conflict and marital adjustmentshowed strong relationships to hormones, accounting for 16% to21 % of the variance in the rates of change of cortisol, ACTH, andnorepinephrine measured serially across the interaction. In con-trast, husbands' endocrine data was not significantly related tonegative behavior or marital quality.

Across these studies, the relationships between physiologicalchange and negative behaviors have typically been stronger forwomen than for men, and women's physiological changes follow-ing marital conflict show greater persistence than men's (Ewart etal., 1991;Fehm-Wolfsdorfetal., 1999; Kiecolt-Glaser etal., 1993,1997; Kiecolt-Glaser, Newton, et al., 1996; Malarkey et al., 1994;Mayne et al., 1997). Moreover, wives' physiological changes weremore closely linked to conflict behavior and marital adjustmentthan husbands' (Ewart et al., 1991; Jacobson et al., 1994; Kiecolt-Glaser et al., 1993, 1997; Kiecolt-Glaser, Newton, et al., 1996;Malarkey et al., 1994; Morell & Apple, 1990).

Might women's greater physiological responsiveness to maritalconflict simply reflect broader patterns of response to acute stres-sors? In fact, the gender differences observed in the marital inter-action studies are particularly noteworthy because men appear toshow larger cortisol and catecholamine responses to a range oflaboratory stressors than do women. For example, comparison ofeffect sizes in response to harassment revealed that cortisol reac-tivity in men was twice that of women (Earle, Linden, & Wein-berg, 1999). Similarly, women's cortisol levels decreased whenanticipating the stress of public speaking, whereas men's cortisolincreased, and although both sexes responded with cortisol in-

creases to speaking and performing mental arithmetic in front of anaudience, men's responses were 1.5- to 2-fold higher than those ofwomen (Kirschbaum, Wust, & Hellhammer, 1992). Moreover, thefact that wives displayed larger blood pressure changes than hus-bands in two studies (Ewart et al., 1991; Mayne et al., 1997), aswell as greater increments in epinephrine in response to conflict(Malarkey et al., 1994), is particularly interesting because mentypically show larger blood pressure and epinephrine increases inresponse to most laboratory stressors than do women, although the"gender relevance" of a stressor may modulate responsivity(Smith, Gallo, Goble, Ngu, & Stark, 1998; Stoney, Davis, &Matthews, 1987). Thus, the sex differences observed during mar-ital conflict are at variance with broader physiological patterns ofresponding to acute stressors.

Alternatively, might women's greater physiological responsive-ness to marital conflict reflect gender differences in physiologicalself-monitoring? For example, if males show greater accuracy thanfemales at detecting physiological signals, men might be moreaware of the relationships between conflict and arousal, and thusmight make greater efforts to decrease this arousal; women, notself-monitoring as carefully, might not decrease their arousal dur-ing conflict, to the detriment of their health. Moreover, womenmight be socialized to regulate themselves based on others' be-havior (Cross & Madson, 1997b). Consistent with this interpreta-tion, Levenson et al. (1994) reported that mood self-ratings madewhile a couple watched the videotape of their conflict session"several days earlier" were correlated with husbands' (but notwives') autonomic responses recorded previously during conflict.As these authors noted, a number of laboratory signal-detectionstudies have demonstrated that males show greater accuracy thanfemales at detecting physiological signals when situational cuessuch as the presence of fear-producing stimuli are experimentallycontrolled, similar to their data (Pennebaker & Roberts, 1992).

However, in naturalistic settings that provide multiple cues(including data from several large field studies), women and menshow equal accuracy at detection of blood pressure, heart rate, orblood glucose (reviewed in Pennebaker & Roberts, 1992), incontrast to laboratory studies such as that of Levenson et al.(1994). Thus, the sex differences observed during marital conflictare at variance with broader physiological patterns of respondingto laboratory stressors, and relevant data do not suggest that thesedifferences are due to gender discrepancies in the ability to useself-monitoring to moderate physiological arousal.

Among the first set of studies in Table 4, only two of those thatanalyzed data by gender reported significantly greater responses inmen than women on any physiological parameter. In the first ofthese, discussion of marital problems produced significant in-creases in blood pressure for both men and women, as well aschanges in a number of other cardiovascular parameters (Broad-well & Light, 1999). Subsequent median splits on a measure offamily support suggested that both men and women who reported

'high family support had lower vascular resistance indexes thanthose low on support across all four tasks. In addition, men whoreported high support had lower blood pressure than those withless support, and these differences were observable across all fourexperimental periods (baseline, discussion of events of the day,conflict, and recovery). However, the fact that these latter analyseswere not limited to perceptions of spousal support makes interpre-tation difficult, particularly in view of the absence of significant


group by experimental period interactions that would suggestgreater specificity for the marital relationship.

One other study found no relationships between behaviorallycoded affect during conflict and cardiovascular, immune, or cor-tisol data among wives; however, although there were no signifi-cant findings for women, there were also none for men, unless thehusband was high on cynical hostility (G. E. Miller, Dopp, Myers,Felten, & Fahey, 1999). Anger displayed during the conflict wasassociated with greater elevations in blood pressure, cortisol, andnatural killer (NK) cell cytotoxicity and numbers among men highin cynical hostility (but not among those who were low). Whymight this study differ from related work?

Methodological differences among the studies in Table 4 pro-vide one explanation. The length of the conflict discussions variedfrom 10 to 40 min, and the strategies (and time allotted) to preparecouples for the task also differed. For example, G. E. Miller et al.(1999) simply asked couples to spend 15 min attempting to resolvean issue previously identified as problematic. In contrast, in arelated set of endocrine-immune studies, the experimenter firsttalked with couples for 10 to 20 min to help identify the best topicsfor discussion among issues previously identified as problematic,and then the conflict discussion lasted 30 min (Kiecolt-Glaser etal., 1993, 1997; Kiecolt-Glaser, Newton, et al., 1996; Malarkey etal., 1994). Similarly, in an immunological project from anotherlaboratory, the researcher sat with couples and facilitated the40-min conflict interactions when they slowed (Mayne et al.,1997).

The length of the problem discussion and the type or intensity ofpreparation are likely to be most critical when the marital adjust-ment of the sample is average or higher; in nondistressed mar-riages, the kinds of negative behaviors that magnify physiologicalchange during conflict occur at lower frequencies (Fincham &Beach, 1999). In addition, as noted previously, unhappy couplesare less likely to volunteer for marital research projects than thosewho are more satisfied with their spouse (Bradbury & Karney,1993); indeed, G. E. Miller et al. (1999) reported that 14 coupleswho had initially responded to their newspaper advertisementdeclined to participate after learning that they would be asked todiscuss a disagreement in their marriage. Thus, demonstratingphysiological consequences of marital problem discussions re-quires particular effort in selecting distressed couples and maxi-mizing the involvement of couples in conflict discussions.

Physiological Reactivity to Neutral or Impersonal SpousalDisagreements

Unlike the participants in the conflict interaction studies inTable 4, the couples studied by Smith and his colleagues discussedimpersonal topics, not relationship issues (Brown & Smith, 1992;Smith & Brown, 1991; Smith et al., 1998). In the first of this series,couples were given an incentive to influence the spouse (Brown &Smith, 1992; Smith & Brown, 1991). Although both spouses weregiven the same inducement, each was led to believe that he or shealone had an incentive. Under these conditions, husbands showedgreater elevations in systolic blood pressure than wives, and theseelevations were correlated with husbands' hostile and controllingbehavior (Brown & Smith, 1992). However, further analyses fromthis project showed that higher cynical hostility in husbands wasassociated with greater blood pressure reactivity in their wives,

whereas wives' hostility was not related to either their own or theirhusbands' reactivity (Smith & Brown, 1991). Among husbands,only their own hostility predicted their blood pressure changes.

Results of a subsequent study suggest that the type of interper-sonal demand might contribute to differences in husbands' andwives' cardiovascular responses to impersonal interactions. In thisstudy, couples in a high-challenge condition were told that theiraudiotaped discussion of two community issues would be rated forverbal competence, whereas low-challenge participants were toldonly to speak audibly and clearly (Smith et al., 1998). In contrast,to manipulate disagreement, the experimenters randomly assignedspouses to either the same or opposing sides for discussion ofcommunity issues. Thus, the authors manipulated disagreement (acommunion stressor) and an achievement challenge (an agencystressor) to assess whether men and women would be differentiallyreactive to stressors that embody gender role characteristics.

As predicted, disagreement elicited heightened heart rates andblood pressure among wives, but not husbands (Smith et al., 1998).In contrast, the achievement challenge had no effects on women'scardiovascular responses, whereas men in the high-achievement-challenge condition displayed larger heart rate and blood pressureincreases than those in the low-challenge condition, consistentwith earlier findings on men's greater response to incentives(Brown & Smith, 1992). Of importance, these differences werealso observed while couples were preparing for the discussion, notjust during the time when they were actually talking.

These data suggest that qualitatively different interpersonal de-mands may differentially activate husbands' and wives' cardio-vascular responses. As reviewed in the prior section, studies inwhich spouses discuss areas of marital disagreement showed thatnegative or hostile behavior during conflict was clearly associatedwith physiological alterations, with larger differences amongwomen than men. The fact that wives respond to spousal disagree-ments, even on impersonal topics, with larger cardiovascular re-sponses than husbands (Smith et al., 1998) is consistent with thosestudies.

Depression

Both syndromal depression and depressive symptoms arestrongly associated with marital distress (Beach et al., 1998; Fin-cham & Beach, 1999). The strength of the tie is sizable—forexample, one study found a 10-fold increase in risk for depressivesymptomatology associated with marital discord (O'Leary, Chris-tian, & Mendell, 1994); similarly, data from a large epidemiolog-ical study demonstrated that unhappy marriages were a potent riskfactor for major depressive disorder for both men and women,associated with a 25-fold increase over untroubled marriages(Weissman, 1987). The relationship appears to be bidirectional,with poor marriages enhancing depressive symptoms, and depres-sion promoting poorer marital quality (Beach et al., 1998; Fincham& Beach, 1999).

The magnitude of the relationship between marital distress anddepressive symptomatology is similar for women and men(O'Leary et al., 1994), a surprising finding because of the evidencethat relationships are more important for women (Cross & Mad-son, 1997b), as well as the fact that women are twice as likely tobe clinically depressed as men (Weissman, 1987). However, recentwork suggests that marital adjustment and depression show differ-


ent relationships for men and women; among a sample of recentlymarried couples who provided marital adjustment and depressivesymptom data at two time points, the path for men emerged fromdepressive symptoms to marital adjustment, whereas data forwomen showed a stronger prospective association in the path fromadjustment to depression (Fincham, Beach, Harold, & Osborne,1997).

Marital disagreements appear to contribute to poorer mentalhealth after considering the contribution of marital quality (Mc-Gonagle & Schilling, 1992), paralleling the evidence from maritalinteraction studies that marital conflict behaviors predict physio-logical change after controlling for marital adjustment (Kiecolt-Glaser et al., 1993, 1997; Kiecolt-Glaser, Newton, et al., 1996).Indeed, day-to-day marital disagreements were better predictors ofdaily mood variation among married individuals than any othercommon stress experiences (Bolger, Delongis, Kessler, & Weth-ington, 1989). In fact, among individuals with remitted depression,expressed emotion and marital adjustment predicted the sameproportion of variance in patients' outcomes (Hooley & Teasdale,1989); however, the single best predictor of relapse was a patient'sresponse to one question: "How critical is your spouse of you?"(p. 229).

The association of marital distress with higher rates of bothsyndromal depression and depressive symptoms has importantimplications for physical functioning. For example, datafrom 11,242 outpatients in the Medical Outcomes Study showedthat patients with either a current depressive disorder or depressivesymptoms in the absence of a syndromal disorder had worsephysical, social, and role function; worse perceived current health;and greater bodily pain than patients with no chronic conditions(Wells et al., 1989). The poorer functioning that was uniquelyassociated with depressive symptoms was comparable to—or evenworse than—that uniquely associated with eight chronic medicalconditions.

Depression alters cardiovascular, immune, and endocrine func-tion, and these alterations are sufficient to enhance a variety ofhealth threats (Classman & Shapiro, 1998; Herrmann et al., 1998;Kiecolt-Glaser, Page, et al., 1998; Penninx, Guralnik, Ferrucci, etal., 1998; Penninx, Guralnik, Pahor, et al., 1998; Simonsick, Wal-lace, Blazer, & Berkman, 1995). In addition to physiologicalalterations, distressed individuals are also more likely to havepoorer health habits including a greater propensity for alcohol anddrug abuse, inadequate sleep and nutrition, and less exercise(Kiecolt-Glaser & Glaser, 1988), all of which have negative healthinfluences in their own right.

Trait Hostility

Hostility refers to a constellation of cognitive, affective, andbehavioral features including interpersonal mistrust and suspi-ciousness and cynicism about human nature, along with tendenciesto experience anger and resentment and to behave uncooperativelyand aggressively (Smith, 1992). Identified as an independent riskfactor for premature mortality and morbidity, particularly coronaryheart disease (T. Q. Miller, Smith, Turner, Guijarro, & Hallet,1996), hostility could contribute to poorer health outcomes byeroding social relationship quality, heightening physiological re-sponses to stressful social interactions, and increasing unhealthy

behaviors such as cigarette smoking and alcohol use (Scherwitz &Rugulies, 1992; Smith, 1992).

Compared with their low-hostile counterparts, individuals whoare high in hostility show more negative conflict behaviors duringmarital interaction, an association that is more prominent amonghostile husbands than among hostile wives (Newton, Kiecolt-Glaser, Glaser, & Malarkey, 1995; Smith, Sanders, & Alexander,1990). Similarly, hostile husbands, but not wives, report lowerlevels of marital satisfaction in cross-sectional studies (Houston &Kelly, 1989; Smith, Pope, Sanders, Allred, & O'Keefe, 1988).Further, in one prospective study of newlywed couples married anaverage of 5 months at initial assessment, hostile husbands showedsignificant linear decreases in marital satisfaction over the first 3years of marriage (Newton & Kiecolt-Glaser, 1995). The wives ofhostile husbands showed similar significant linear decreases inmarital satisfaction. In contrast, wives' trait hostility was notassociated with changes in their own, or their husbands', maritalsatisfaction. Trait hostility also is related to depressive symptom-atology in married couples (Brummett et al., 2000). Hostile hus-bands and wives report higher levels of depressive symptoms, and,for wives, being married to a hostile spouse adds to depression.

Trait hostility contributes to negative marital interaction anderodes marital quality, and it also heightens the physiologicalconsequences of marital conflict. It is well documented that hostileindividuals show cardiovascular hyperreactivity in interpersonalcontexts characterized by harassment or provocation, presumablybecause their psychological characteristics predispose them toperceive, and react strongly to, interpersonal threats (Suls & Wan,1993). Predictably, trait hostility heightens blood pressure andheart rate responses to impersonal marital discussions (Smith &Brown, 1991; Smith & Gallo, 1999) and also amplifies cardiovas-cular and cortisol responses that accompany angry marital conflict(G. E. Miller et al., 1999). These responses are uniformly apparentfor hostile husbands. In contrast, hostile wives do not exhibitheightened physiological reactivity during marital conflict, butwives in general show heightened systolic blood pressure wheninteracting with hostile husbands (Smith & Brown, 1991).

In sum, hostile characteristics are positively associated withmarital dissatisfaction and conflict behaviors, and several studiessuggest they contribute to heightened physiological reactivity dur-ing marital discussions and conflict interactions. Preliminary evi-dence shows that hostility contributes to depression in marriedcouples. There is a fairly uniform gender disparity across theseobservations, with hostile husbands, more so than hostile wives,showing greater dissatisfaction, conflict, and hyperreactivity; incontrast, wives' social, emotional, and physiological functioningare all negatively affected by their husbands' trait hostility.

Health Habits

Supportive relationships can directly influence health by facil-itating health-promoting behaviors and decreasing maladaptivecoping behaviors (Lewis, Rook, & Schwarzer, 1994). For example,higher marital adjustment was associated with better compliancewith a blood pressure medication regimen (Trevino, Young, Groff,& Jono, 1990). Among women who underwent gastric restrictionsurgery for the treatment of morbid obesity, marital dissatisfactionwas associated with weight gain a year later (Hafner, Rogers, &Watts, 1990). In longitudinal data from men, positive marital


interaction (assessed by both self-report and behavioral indices)reduced the probability of risky health habits (poor eating habits,substance use, and inadequate sleep); moreover, risky lifestylescontributed to poorer health after controlling for initial healthstatus (Wickrama, Conger, & Lorenz, 1995).

Some of the strongest data indicate that marital conflict may beboth a precursor and a consequence of alcohol and drug abuse(O'Farrell, Hooley, Fals-Stewart, & Cutter, 1998). Among indi-viduals who underwent behavioral marital therapy (BMT) foralcoholism, patients with high-EE spouses were more likely torelapse, had a shorter time to relapse, and drank on a greaterpercentage of days in the year after starting BMT than patientswith low-EE spouses (O'Farrell et al., 1998). A longitudinal studyassessed the frequency of negative consequences from alcohol use(Horwitz & White, 1991); after controlling for alcohol problems atbaseline, participants' responses to a single question about howoften serious conflicts arose within the marriage were associatedwith problem drinking for men, but not women, 3 years later.Similarly, among middle-aged couples, unhappy husbands con-sumed more alcohol than happy husbands, with no differences forwives (Levenson et al., 1993). In contrast to the results of thesetwo studies, when predisposing factors and mental health statuswere statistically controlled, women who reported greater maritalconflict were more likely to smoke and to drink a moderate toheavy amount of alcohol; marital stress was not associated withhealth practices among men (Cohen, Schwartz, Bromet, & Parkin-son, 1991). Recent spousal conflict was associated with greatertranquilizer use for both genders in a nationwide sample of Finnishmen and women (Appelberg, Romanov, Honkasalo, & Kosken-vuo, 1993). Data such as these suggest that interpersonal conflictin close relationships intensifies maladaptive health practices.Could health habits and/or depression-related marital distress ac-count for the physiological differences observed in maritalinteraction?

This issue was addressed in a series of articles that reported datafrom newlywed couples who were selected on the basis of ex-tremely stringent mental and physical health criteria (Kiecolt-Glaser et al., 1993; Kiecolt-Glaser, Newton, et al., 1996; Malarkeyet al., 1994). Couples were eliminated from consideration duringan initial phone interview if either spouse reported a previousmarriage or children, any acute or chronic health problems thatmight have immunological or endocrinologic consequences, ifthey took any medications except birth control pills, if they drankmore than 10 alcoholic drinks per week or used any street drugs,if they smoked, if they used caffeine excessively, or if they werenot within 20% of their ideal weight for their height. A second setof phone interviews assessed both current and lifetime psychiatricdisorder data as well as a detailed medical history. Participantswere excluded who had met Diagnostic and Statistical Manual ofMental Disorders (3rd ed., rev.; American Psychiatric Association,1987) criteria for any psychotic diagnosis, any depressive or anx-iety disorder other than simple phobia, or substance abuse. Amongparticipants selected on the basis of these stringent criteria, thosewho exhibited more hostile or negative behaviors during a maritalproblem discussion showed greater immunological, endocrino-logic, and cardiovascular alterations relative to low-negative-behavior participants.'

Subsequent analyses of couples who were high or low on hostilebehavior during conflict showed no differences on age, education,

income, social support in other relationships, social desirability,negative or positive mood, depressive or anxiety symptoms, AxisII personality dimensions, parental history of hypertension or heartdisease, length of marriage or how long they had dated beforemarriage, or health habits. Baseline heart rate and blood pressuredid not diverge before the conflict discussion, and the two groupsdid not differ in cardiovascular reactivity to a mental arithmetictask in the afternoon; this latter similarity between groups wasespecially important in light of data linking cardiovascular reac-tivity with short-term endocrinologic and immunological changes(Kiecolt-Glaser, Cacioppo, Malarkey, & Glaser, 1992). Admissionto a hospital research unit provided a uniform environment, simul-taneously controlling factors such as physical activity, diet, andcaffeine intake (Kiecolt-Glaser & Glaser, 1988). Accordingly, itseems reasonable to conclude that the behavior of the couplesduring their 24 hr together, rather than extraneous factors, pro-duced the observed cardiovascular, endocrine, and immunologicalchanges.

On the basis of the evidence available at the time, Burman andMargolin (1992) concluded that "marital variables affect health,but the effect is indirect and nonspecific" (p. 39). Although maritalfunctioning undoubtedly influences health indirectly by meansof modulation of mood and health habits, the marital interac-tion studies published in the decade since their review havedemonstrated direct and specific influences on key physiologicalmechanisms.

Health Consequences: Linking Physiological Changes toMorbidity and Mortality

Burman and Margolin (1992) argued that the most convincingway to document a causal relationship between marital functioningand health status would be first to confirm that marital interactionhad direct effects on physiological processes and then to show thatindividuals who exhibited physiological changes were more likelyto develop health problems; this is certainly the most stringent test,and none of the studies reviewed here have all of these compo-nents. Indeed, the links between autonomic, endocrine, and im-mune responses to challenge and health are tenuous, with re-sponses to laboratory challenge typically returning to resting levelswithin 1 to 2 hr after cessation of the stressor (Kiecolt-Glaser et al.,1992). Although data from some of the marital interaction studiesare consistent with evidence from this "challenge" paradigm, oth-ers are particularly provocative precisely because they deviatefrom this pattern—for example, among newlyweds immunologicaldifferences endured for 22 hr across 10 different assays after amarital conflict, consistent with persistent endocrinologic alter-ations (Kiecolt-Glaser et al., 1993; Kiecolt-Glaser, Newton, et al.,1996). Moreover, although direct effects on etiology have yet to bedemonstrated, marital functioning unquestionably has consequen-tial influences on symptom expression (a key component of dis-ability). In addition, the marital interaction literature may be ex-amined with an alternative question: Are the physiologicalalterations that have been demonstrated to date large enough tohave clinical significance? Evidence from each of these vantagepoints is evaluated below.

The best evidence relating marital functioning to health statuscomes from illnesses that have immunological or cardiovascularcomponents or mediators. The endocrine system serves as one

492 K1ECOLT-GLASER AND NEWTON

important gateway for both; stressors can provoke the release ofpituitary and adrenal hormones that have multiple effects, includ-ing alterations in cardiovascular and immune function (Dhabhar &McEwen, 1997; Glaser & Kiecolt-Glaser, 1994; Kuhn, 1989). Theendocrine system's involvement in the pathogenesis of stress-related disease processes is probably mediated through frequentsmall daily excursions in hormonal levels following stressfulevents. For example, chronic stimulation of cortisol and catechol-amine secretion at lower levels has been linked to cardiovascularpathology (Kuhn, 1989) and immunological dysregulation (Dhab-har & McEwen, 1997; Glaser & Kiecolt-Glaser, 1994). Cortisolfacilitates the vasoconstrictive effects of catecholamines; accord-ingly, the combination of catecholamine and cortisol responses isimportant for pathogenesis in cardiovascular disease (Fredrikson,Tuomisto, & Bergman-Losman, 1991). Furthermore, both cat-echolamines and cortisol are associated with immunological dys-regulation (Glaser & Kiecolt-Glaser, 1994). The ability to unwindafter stressful encounters (i.e., quicker return to one's neuroendo-crine baseline) influences the total burden that stressors place onan individual (Frankenhaeuser, 1986). Stressors that are resistantto behavioral coping, particularly stressors perceived as unpredict-able and uncontrollable, may continue to be associated with ele-vated stress hormones even after repeated exposure (Baum, Cohen,& Hall, 1993).

Dhabhar and McEwen (1997) speculated that chronic stressdysregulates immune function in part by means of disruption ofcircadian rhythms, leading to elevated cortisol at the time of daywhen levels are normally low. Newlyweds' cortisol data are pro-vocative in this context. Hours after a laboratory conflict task hadended, negative behaviors accounted for 24% of the variance in"daytime" cortisol, that is, pooled samples collected hourly from8 a.m. through 10 p.m. (Kiecolt-Glaser, Newton, et al., 1996). Asan illustration of the biological implications of these data, a wifewho fell 1.5 standard deviations above the mean for the demand-withdraw behavioral sequence and at the mean on all other pre-dictors (i.e., holding other values constant) would show a predictedcortisol level of 15.28 ng/ml, whereas a woman 1.5 standarddeviations below the mean would have a predicted cortisol level ofonly 7.71 ng/ml—about half as large. Similarly, among olderwomen in long-term marriages, multilevel modeling of cortisolchange showed a mean slope of .04 and variance of .70 in thoseslopes. A model for prediction of that variation in slopes showedthat conflict behavior accounted for 21% of the variance in the ratethat cortisol changed when measured serially across the problemresolution task (Kiecolt-Glaser et al., 1997). These data are likelyto underestimate the actual physiological impact of marital dis-cord, because both the younger and older couples were generallyquite happy; only 3% of newlyweds and 13% of the older samplehad marital adjustment scores in the distressed range. Thus, neg-ative or hostile marital conflict behaviors were linked to biologi-cally relevant elevations in cortisol even in relatively nondistressedcouples.

For what kinds of health outcomes might such changes haverelevance? Recent studies have demonstrated substantial relation-ships between stress and wound healing, a process in which theimmune system plays a key role (Kiecolt-Glaser, Marucha, Ma-larkey, Mercado, & Glaser, 1995; Kiecolt-Glaser, Page, et al.,1998). Even a transient, commonplace stressor, academic exami-nations, can substantially delay wound repair: Oral wounds

placed 3 days before an exam healed an average of 40% moreslowly than those made in the same individuals during summervacation (Marucha, Kiecolt-Glaser, & Favagehi, 1998). Stress-induced elevations in cortisol can alter the carefully regulateddynamic system that controls development of the immune re-sponse at the wound site, suppressing production of cytokines, keymessenger molecules that help cells communicate (DeRijk et al.,1997; Hubner et al., 1996). Indeed, in a study that measuredimmune function at the site of blister wounds, low cytokine pro-ducers reported more stress and more negative mood than highproducers, and the former also had higher levels of salivary cor-tisol (Glaser et al., 1999). Consistent with these data, male coro-nary artery bypass patients who received greater spousal supportused less pain medication, had a more rapid discharge from thesurgical intensive care unit, and spent fewer total days in thehospital (Kulik & Mahler, 1989).

Six studies have demonstrated an association between maritalquality or marital interaction and immune function (Kiecolt-Glaseret al., 1987, 1988, 1993, 1997; Mayne et al., 1997; G. E. Miller etal., 1999). A central question throughout much of the psychoneu-roimmunology literature has been the extent to which stress-induced immune changes have consequences for morbidity andmortality (Kiecolt-Glaser, 1999). The data on wound repair pro-vide clear evidence of important health effects in one domain. Inaddition, several studies have demonstrated stress-related modula-tion of vaccine responses in both younger and older participants,including alterations in response to a mild stressor, academicexams (Glaser, Kiecolt-Glaser, Bonneau, Malarkey, & Hughes,1992; Kiecolt-Glaser, Glaser, et al., 1996; Vedhara et al., 1999).Vaccine responses provide an excellent proxy for infectious illnessrisk because they demonstrate clinically relevant alterations inimmune responses to challenge under well-controlled conditions;adults who show poorer responses to vaccines and other antigenicchallenges also experience higher rates of clinical illness (E. A.Burns & Goodwin, 1990; Hobson, Curry, & Beare, 1972; Patri-arca, 1994). Thus, these data provide a window on the body'sresponse to other pathogens. In accord with these findings, indi-viduals who reported enduring interpersonal difficulties with fam-ily or friends were significantly more likely to develop a coldfollowing inoculation with a cold virus than those who were notexperiencing interpersonal strains (Cohen et al., 1998). Thus, thereis good evidence (albeit indirect) to suggest that the immune andendocrine dysregulation associated with marital discord is likely tobe consequential for two broad health outcomes, wound healingand infectious disease risk.

Stress-related immune dysregulation is also a catalyst for clin-ical change in rheumatoid arthritis, an autoimmune disease. Usinga prospective design, Zautra and colleagues (Zautra et al., 1998)linked interpersonal stress with changes in clinician-rated diseaseactivity as well as immune function, a notable addition to therheumatoid arthritis literature; in addition, they found that womenwith better spousal relationships appeared to be protected—that is,they did not show increases in disease activity. In related workwith patients with rheumatoid arthritis, the immune-stimulatinghormones prolactin and estradiol were significantly positively cor-related with interpersonal conflicts, depression, coping inefficacy,and clinician ratings of disease activity; for Systemic autoimmunediseases such as rheumatoid arthritis, increases in aspects of im-mune function are maladaptive, because they are associated with


disease flare-ups (Zautra, Burleson, Matt, Roth, & Burrows, 1994).Thus, interpersonal stress was linked to both endocrine and im-mune alterations, and these changes were also associated withclinician-rated disease activity as well as self-reported joint ten-derness in well-designed prospective studies. Similarly, other datafrom patients with rheumatoid arthritis showed that negativespouse behavior predicted poorer pain outcomes longitudinallywith controls for baseline pain (Waltz et al., 1998). The potentiallong-term significance of such changes is suggested by data froma community cohort of patients with rheumatoid arthritis who wereassessed every 6 months for up to 9.5 years; married patients hada slower progression of functional disability than the unmarried(Ward & Leigh, 1993).

Considerable research on cardiovascular disease has beenguided by the reactivity hypothesis, the premise that excessivecardiovascular reactivity to stress is a risk factor for the develop-ment of hypertension and cardiovascular disease, particularly ifresponses occur relatively frequently and at high intensity (Carelset al., 1998; Smith et al., 1998). In the studies reviewed above,marital conflict was reliably associated with heightened bloodpressure and heart rates (Broadwell & Light, 1999; Ewart et al.,1991; Flor et al., 1995; Kiecolt-Glaser et al., 1993; Mayne et al.,1997; Morell & Apple, 1990; Schwartz et al., 1994; Thomsen &Gilbert, 1998). Of importance, among a sample of 43 patients withessential hypertension, a 10-min marital problem-solving task pro-duced clinically significant increases in blood pressure, with par-ticipants reaching a mean of 160/100 mmHg. Twenty of thesecouples subsequently completed marital communications training,and their blood pressure reactivity to marital arguments was re-duced compared with those who did not undergo training (Ewart,Taylor, Kraemer, & Agras, 1984), a notable demonstration of thelinkage between marital factors and clinically relevant physiolog-ical responses.

When a marriage becomes sufficiently abrasive, a bona fidedisagreement is not essential to promote heightened cardiovascularresponses; marital conflict recall in the absence of the spouse wassufficient to elevate blood pressure among women low on maritaladjustment (Carels et al., 1998). More spousal contact was asso-ciated with elevated evening blood pressure among individualswith mild hypertension who reported lower marital cohesion(Baker et al., 1999). In addition to these direct pathways, there arealso important indirect routes; higher marital adjustment was as-sociated with better compliance with a blood pressure medicationregimen (Trevino et al., 1990), and depressive affect has beenlinked to cardiovascular morbidity and mortality (Glassman &Shapiro, 1998; Simonsick et al., 1995). A meta-analysis suggestedthat familial sources of social support are associated with reliablepositive effects on blood pressure regulation (Uchino, Cacioppo, &Kiecolt-Glaser, 1996). Finally, lack of disclosure to one's spousepredicted poorer recovery on three indices 1 year after a myocar-dial infarction: rehospitalization and/or death, post-myocardial-infarction chest pain, and perceived health (Helgeson, 1991). Thus,marital functioning has obvious negative influences on hyperten-sion and cardiovascular disease through both direct and indirectroutes.

Marital interaction unquestionably alters symptom expression ina number of chronic conditions. Spousal reinforcement of painbehaviors has been associated with subsequent illness chronicity(Paulsen & Altmaier, 1995; Romano et al., 1995; Turk et al.,

1992); hence, the finding that patients who reported increasedmarital adjustment following a spouse-assisted coping skills inter-vention were more likely to report less physical disability andshow less pain behavior is particularly provocative (Keefe et al.,1996). In this context, the evidence that women may be at greaterrisk for pain-related disability than men (Unruh, 1996) may havebeen a factor in a large longitudinal study that found greaterphysician-certified disability related to marital conflict for women,but not men (Appelberg et al., 1996).

Two large prospective epidemiological studies implicated mar-ital strain as a factor in the development of ulcers (Levenstein etal., 1999; Medalie, Stange, Zyanski, & Goldbourt, 1992), findingsthat echo earlier cross-sectional work (Alp, Court, & Grant, 1970;Gillies & Skyring, 1968). Although the data were based on self-reports, one of the studies assessed the accuracy of participants'recall by requesting medical records for 25% of all ulcer cases(Medalie et al., 1992). Definite confirmatory evidence was foundin 77% of these cases; false positives included patients who werebeing treated by their physician with an ulcer diet despite negativeradiological findings and cases in which the radiologist examiningthe radiographic data disagreed with the original interpretation.Endocrine and immune dysregulation play a role in addition tohealth habits (smoking, alcohol use, and lack of sleep), and stressalso alters key gastrointestinal processes such as blood flow andgastric acid secretion (Levenstein et al., 1999). Thus, maritalfunctioning could play a role in ulcers' multifactorial origin andcourse through a number of routes.

If abrasive relationships provoke larger and more frequent im-munological, endocrinologic, and cardiovascular changes, thenindividuals in troubled relationships could be at greater risk for avariety of health problems over time. Distressed families experi-ence roughly twice as many tensions per day as nondistressedfamilies (Christensen & Margolin, 1988; Margolin, Christensen, &John, 1996). There is also greater spillover of conflict from onetopic to another and greater "contagion" between marital andchild-related tensions among unhappy couples than those who aremore satisfied (Margolin et al., 1996). Moreover, distressed cou-ples are more likely to experience continuance of tensions, partic-ularly those that repeat in ritualized patterns at the same time onsubsequent days (Margolin et al., 1996). In contrast to stressorswithout an interpersonal component, those that involve conflicthave an increasing emotional impact as stressors occur over days,and they account for a large portion of the variance in daily mood(Bolger, Delongis, Kessler, & Schilling, 1989). Older adults arelikely to be at greatest risk, related to the greater physical vulner-ability that accompanies aging. For example, age and distressappear to interact to promote immune regulation: Older adultsshow greater immunological impairments related to stress or de-pression than young adults (Kiecolt-Glaser, Glaser, et al., 1996;Schleifer, Keller, Bond, Cohen, & Stein, 1989). Accordingly,troubled marriages are likely to be more consequential for thehealth of older couples.

Conceptual Perspectives on Gender Differences

The physiological studies of marital interaction from the pastdecade provide convergent evidence that gender is an importantmoderator of the pathway from negative marital conflict behaviorsto physiological functioning; this pathway is stronger for women


than for men, and women's physiological changes following mar-ital conflict show greater persistence than men's (Ewart et al,1991; Jacobson et al., 1994; Kiecolt-Glaser et al., 1993, 1997;Kiecolt-Glaser, Newton, et al., 1996; Malarkey et al., 1994; Mayneet al., 1997). As noted earlier, this gender pattern stands in contrastto broader patterns of responding to acute stressors in which menshow larger physiological responses than women (Earle et al.,1999; Kirschbaum et al., 1992). In this section we consider evi-dence that a triad of loosely integrated gender-linked factors—self-representations, traits, and roles—contributes to gender differ-ences in the pathways leading from the marital relationship tophysiological functioning and health outcomes. This perspectivefurnishes a preliminary working model of the origins of thesegender differences, and provides a framework to guide research onmarital functioning, gender, and health.

It is generally accepted that gender is a multidimensional con-struct composed of biological, cognitive, affective, and behavioralfeatures that are loosely integrated (Ashmore, 1990). Self-representations have emerged as one key dimension of gender.Self-representations, or self-construals, are central to social infor-mation processing. They guide social cognition by directing atten-tion to information that is emotionally salient to the self and byfacilitating recall of self-relevant life experiences. Moreover, theyserve motivational and regulatory functions, promoting, or some-times hampering, short- and long-term efforts to achieve valuedgoals (Cross & Madson, 1997b).

Recent theoretical and empirical developments suggest there arefundamental differences in the content and structure of men's andwomen's self-representations, perhaps as a result of socializationand cultural influences (Cross & Madson, 1997a, 1997b). Specif-ically, women's self-construals, compared with men's, are char-acterized by relational interdependence (Acitelli & Young, 1996;Cross & Madson, 1997a, 1997b). Self-construals characterized byrelational interdependence incorporate representations of close andsignificant others (e.g., a spouse), so that self-attributes, charac-teristics, and preferences are represented within the context ofclose, often dyadic, relationships. Individual goals, strivings, andregulatory functions are influenced by valued personal relation-ships that are represented hand-in-hand with the self. In marriage,wives' interdependent self-systems will include representations oftheir husbands, and their thoughts and feelings in marriage will bepartially regulated by and responsive to not only their own behav-ior, but also that of their husbands. In contrast, men's self-construals are less relational than are women's. Although menclearly have, value, and strive for close relationships, havingself-construals that are less relational means that their social in-formation processing and self-regulation are less influenced byclose, dyadic relationships (Cross & Madson, 1997a, 1997b).Instead, evidence suggests that men's self-construals are charac-terized by collective interdependence, or representations of groupmemberships and affiliations that occur at a broader social sphere(Baumeister & Sommer, 1997; Gabriel & Gardner, 1999).

This perspective predicts that wives, by virtue of their morerelationally interdependent self-representations, should be moreattuned to (and less insulated from) the emotional quality ofmarital interactions compared with husbands. Extensive empiricalresearch on gender, emotion, and social relationships is consistentwith this notion. Wives function as the "barometers" of distressedmarriages (Floyd & Markman, 1983), in part because women are

more sensitive to negative marital interactions than are men. Wivesare better than husbands at interpreting their spouse's emotionalmessages (Noller & Fitzpatrick, 1990); distressed wives can moreaccurately decode their husbands' negative messages than thereverse (Notarius, Benson, Sloane, Vanzetti, & Hornyak, 1989).Maladaptive attributions for events in the marital relationship areassociated with less positive behavior and more negative behavioramong wives, whereas husbands' attributions and behavior areunrelated (Bradbury, Beach, Fincham, & Nelson, 1996). Womenare more adversely affected than men by overt expressions ofhostility in marital interactions (Gaelick, Bodenhausen, & Wyer,1985). In the emotional transmission literature, several studieshave provided evidence that husbands' negative emotions predictwives' negative emotions more reliably than the converse (No-tarius & Johnson, 1982; Roberts & Krokoff, 1990), particularlyamong distressed couples (Larson & Almeida, 1999). The fact thatnegative emotions are more contagious than positive emotionsmay lead to increased vulnerability in wives, who are more likelyto be receivers than senders (Larson & Almeida, 1999).

Women also report that they reminisce more frequently aboutimportant relationship events and spend more time thinking abouttheir marital relationships than do men (Burnett, 1987; M. Ross &Holmberg, 1990). Arguments with the spouse are more upsettingto women than to men (Almeida & Kessler, 1998; Bolger, Delon-gis, Kessler, & Wethington, 1989), and wives demonstrate moredetailed and vivid memories of marital disagreements than do theirhusbands (M. Ross & Holmberg, 1990).

These data are consistent in showing that women are moreattuned than men to the emotional quality of marital functioning.This perspective suggests that the observed differences in hus-bands' and wives' physiological responses to negative aspects ofmarital interaction may reflect a biological consequence of gen-dered self-processes; a relational self-construal may increase psy-chological and physiological vulnerability to abrasive interper-sonal relationships. For example, because memories of stressfulexperiences can themselves continue to evoke stress-related phys-iological changes (Baum et al., 1993; Carels et al., 1998), women'sstronger and more enduring memories of marital disagreements arelikely to sustain maladaptive physiological changes such as height-ened cardiovascular responses and elevated stress hormones. Incontrast, husbands' less relational self-representations should in-sulate them from the psychological and physiological conse-quences of marital conflict.

Wives' greater cognitive and emotional sensitivity to maritaldistress and their associated physiological arousal may also be tiedto their greater propensity to mend or end their marriages. Wivesare more likely to voice their discontent with their marriages, andto do so earlier than their husbands (Hagestad & Smyer, 1982;Harvey, Wells, & Alvarez, 1978). Longitudinal data indicate thatwives' autonomic and immunological responses to conflict mayhave greater predictive power than husbands' responses for assess-ing risk for marital discord and dissolution (Gottman & Levenson,1992; Kiecolt-Glaser, Glaser, Cacioppo, & Malarkey, 1998). Inaddition, precisely because they strive for close relationships withothers, wives may be more likely to end a troubled marriage whenthey see their efforts at reparation as failed (Cross & Madson,1997a). Only a quarter to a third of marital separations are directlyprompted by the husband's decision (Kitson, 1982).


A corollary prediction of this conceptual perspective is thatrelational interdependence will render wives more responsive thanhusbands to the positive aspects of marital interactions, not just tothe abrasive qualities (Cross & Madson, 1997b). In contrast, hus-bands' less relational self-representations should provide themwith fewer benefits of positive marital interactions (Cross & Mad-son, 1997b). Although the physiological consequences of positivemarital interactions have not been adequately tested for either menor women, the results of one study reviewed above are consistentwith this proposition (Kiecolt-Glaser, Newton, et al., 1996); wives,but not husbands, who showed higher frequencies of positivebehaviors during marital conflict had lower epinephrine levelsthroughout the day.

A recent trait-based approach to gender differences draws onreminiscent themes and has received preliminary empirical supportin the health arena (Helgeson, 1994). According to this trait-basedmodel, women, compared with men, are more characterized bycommunion, a trait that motivates attention to and focus on others.In contrast, men, compared with women, are more characterizedby agency, a personality trait that motivates separating from othersand focusing on the self. This model posits that communionreliably increases vulnerability to relationship stressors when itoccurs in its extreme form and is unmitigated by agency—that is,when an individual is characterized by a focus on others to theexclusion of the self. In fact, there are supportive data; unmitigatedcommunion was correlated with stronger adverse effects fromstressful interpersonal events, and these stressors accounted for theassociations between unmitigated communion and both psycho-logical distress and poorer metabolic control among adolescentswith diabetes (Helgeson & Fritz, 1996).

Self-representations and the personality traits of agency andcommunion may also affect indirect pathways leading from mar-ital functioning to health outcomes by increasing vulnerability todepression and compromised health habits. Women's greater sen-sitivity to relationship events has been implicated in their height-ened risk for depression (Cross & Madson, 1997b). Similarly, inthe context of social challenges such as those posed by marriage,risk for depression may be increased by the communal personalityorientation that so often characterizes women, whereas agenticpersonality characteristics may decrease the risk of depression(Nolen-Hoeksema & Girgus, 1994). Among couples adjusting to afirst coronary event in the husband, the most distressed wives werethose who were high in unmitigated communion with spouses highin unmitigated agency (Helgeson, 1993). However, in its lessextreme form, communion's positive focus on relationships has anumber of beneficial features, including greater marital satisfac-tion for both the spouse and partner (Bradbury & Fincham, 1988).With regard to health habits, it has been proposed that the extremeor unmitigated forms of communion and agency both contribute topoorer self-care (Helgeson, 1994). Individuals characterized byunmitigated communion, the extreme focus on others that is morelikely to occur among women than men, may prioritize others'needs above their own, thereby neglecting their own health habits.Individuals characterized by unmitigated agency, an extreme focuson the self that is more likely to occur among men than women,may not reliably seek health care because help seeking is viewedas inconsistent with autonomy (Helgeson, 1994).

In sum, convergent evidence suggests that observed genderdifferences in the pathway from negative marital interactions to

physiological responses could be partially driven by self-processesand traits that increase women's vulnerability to abrasive maritalinteractions. In addition, these and other gender-linked self-processes and traits have implications for indirect pathways in-volving depression and health habits and for the impact of positivemarital functioning. Although direct tests are needed, this perspec-tive provides a viable and testable starting point from which toexamine gender differences in pathways from marital functioningto physical health outcomes. Moreover, because marriage occupiesa privileged place among relationships that affect well-being(Glenn & Weaver, 1981), gender-linked traits and self-processesthat operate within marriage may be particularly salient to healthoutcomes. Social conventions and gender-typed marital roles (e.g.,"homemaker," "breadwinner") that continue to characterize manymarriages may buttress the gender typing of self-processes (Cross& Madson, 1997b).

Gender differences in stress exposure that occur within thecontext of marital roles may also contribute to pathways leadingfrom marital functioning to health outcomes. Marital roles refer tothe concrete ways in which time and daily behavior are structuredwithin the context of marriage. In contemporary marriage, partic-ularly salient roles are those that revolve around contributions topaid employment and performance of domestic chores (Glass &Fujimoto, 1994). After controlling for prior reports of mentaldistress, Bird (1999) found that among men and women employedoutside the home, contributing more than 46% of the total domes-tic labor accelerated increases in depressive symptoms. On aver-age, wives contribute up to 53% beyond this threshold, whereashusbands' contributions fall below this threshold; normative ineq-uities in marital roles may contribute to depressive symptomsamong employed wives. The chronic strain of domestic chores alsocontributes to women's tendencies to ruminate or dwell passivelyon their negative emotional states, a response style that increasesvulnerability for later depressive symptoms and disorders (Nolen-Hoeksema & Girgus, 1994; Nolen-Hoeksema, Larson, & Grayson,1999). In reciprocal feedback fashion, rumination in turn has beenlinked to higher levels of domestic role strain, perhaps by damp-ening motivation and vitality to overcome negative situations(Nolen-Hoeksema et al., 1999).

Participation in domestic labor after a day at work also has beenproposed to impair working women's ability to physiologically"unwind" in the evening hours (Frankenhauser, Lundberg,Fredrikson, Melin, & Tuomisto, 1989); this process has potentiallypowerful implications for the disruption of biological circadianrhythms, a proposed mediator of the health consequences ofchronic stress, discussed above. Although more data are needed todirectly test this proposition, indirect indicators of women's par-ticipation in domestic labor (i.e., marital and parental status) havebeen linked with physiological parameters assessed during work-days, nonworkdays, and sleep. For example, compared with un-married female nurses, married nurses showed significantly higherlevels of nighttime urinary cortisol, and their urinary norepineph-rine levels showed no reduction from workdays to days off (Gold-stein, Shapiro, Chicz-DeMet, & Guthrie, 1999). In addition, al-though all nurses showed reductions in heart rate from daytimeworking hours to evening hours, those with children living at homeshowed significantly smaller reductions than did nurses withoutchildren at home (Goldstein et al., 1999). Similarly, compared withworking women without children at home, those with children at


home showed more pronounced elevations in 24-hr cortisol excre-tion during the workday (Luecken et al., 1997), and number ofchildren living at home was positively associated with workingwomen's diastolic blood pressure levels during sleep (James,Gates, Pickering, & Laragh, 1989). In one study that assessedfamily responsibilities, working women's aggregate 24-hr bloodpressure levels were highest among women who had obtained acollege degree and also reported high levels of family responsi-bility (i.e., having more and younger children, and performing agreater percentage of domestic work; Brisson, Laflamme, Moisan,Milot, Masse, & Vezina, 1999). On the other hand, in one study ofmen and women schoolteachers, individuals with children showedgreater reductions in systolic and diastolic blood pressure from theworkday to the evening compared with individuals without chil-dren (Steptoe, Lundwall, & Cropley, 2000).

With one exception, the indirect evidence provided by thesestudies is consistent with the notion that stress exposure accom-panying marital and parental roles contributes to alterations inwomen's circadian physiological levels, including degree of phys-iological unwinding from workday to nighttime and from work-days to nonworkdays. Future research should include direct mea-surements of role stress and should examine the intra- andinterpersonal processes involved in evening and nonworkdayphysiological unwinding. Reciprocal feedback cycles involvingdomestic role strain and rumination may be particularly problem-atic intrapersonal processes for women (Nolen-Hoeksema et al.,1999); likely interpersonal contributors to degree of physiologicalunwinding include the quality of spouses' marital interactionsduring evenings and nonworkdays (Baker et al., 1999; Repetti,1989). In addition, because the role stress of dual-earner marriageshas consequences for husbands, too, studies should include hus-bands and wives. For example, the spillover of interpersonalconflict from workplace to home, and from home to workplace, isstronger among husbands than wives (Bolger, DeLongis, Kessler,& Wethington, 1989); the physiological correlates of such spill-over await investigation.

Overall, what general conclusions can be drawn about gender,physical health, and marital functioning? As described in theintroduction, married people have reliably better physical healthprofiles than unmarried people. At the same time, epidemiologicalstudies of marital status show that being married, as compared withbeing unmarried, is more beneficial for men's health than women's(House et al., 1988; Umberson, 1992). Also, physiological studiesof marital interaction from the past decade provide convergentevidence that marital conflict is likely to have a greater negativeimpact on the health of women than men. In contrast to these cleargender patterns, in the studies reviewed that did not assess phys-iological changes during marital interactions, we did not observesuch strong or uniform gender patterns in associations betweenmarital functioning and physical health outcomes. What mightaccount for the discrepancy? The explanatory power of behavioraldata may be one key factor; within the few marital interactionstudies that made explicit contrasts, objectively measured behav-iors explained considerably more of the variance than self-reportdata among women compared with men (Ewart et al., 1991;Kiecolt-Glaser et al., 1993, 1997; Kiecolt-Glaser, Newton, et al.,1996). In addition, the health outcome studies as a whole assessedmany different aspects of marital functioning (not only maritalconflict per se, which presumably might have revealed consistently

poorer health outcomes for wives), and most did not assess orcontrol for factors such as depression and health habits. Thus, thehealth outcome studies reviewed here reflect the contribution ofmultiple aspects of marital functioning (positive and negative), andmultiple direct and indirect pathways, each with costs that aredifferentially weighted by a host of gender-linked factors. Theconceptual models discussed in this section provide a starting pointfor a more focused and comprehensive exploration of the differ-ential costs and benefits the marital relationship offers men andwomen (House et al., 1988).

Recommendations for the Next Decadeof Research on Marriage and Health

To maximize the chances of demonstrating linkages betweenmarital functioning and health, how should researchers assessmarital relationships? Behavioral data appear to enhance predic-tion (Ewart et al., 1991; Kiecolt-Glaser et al., 1993, 1997; Kiecolt-Glaser, Newton, et al., 1996); in addition, the importance ofassessing both positive and negative aspects of relationships ishighlighted by data from the Whitehall II study, which examinedthe relative impact of three types of support provided by the personthat the respondent named as his or her closest relationship (Stans-feld et al., 1998). When researchers contrasted the effects ofconfiding-emotional support, practical support, and negative as-pects of the relationship (negative interaction and conflict), theyfound that the negative effects of close relationships were inde-pendent and powerful predictors of poor health functioning (bothphysical and psychological), after adjustment for age, employmentgrade, baseline ill health, and negative affect. Moreover, theseeffects did not appear to be mediated through health practices.These data suggested that negative aspects of close relationshipshave an etiological role that is independent of baseline illness.Although the relationships assessed were not exclusively marital,the findings are notable because they demonstrate that conflictsand negative interactions are more than simply the absence ofsupport (Berkman, 1998).

Within the few marital interaction studies that have made ex-plicit contrasts, negative behavior appears to be more closely tiedto physiological changes than positive behavior (Ewart et al.,1991; Kiecolt-Glaser et al., 1993, 1997; Kiecolt-Glaser, Newton, etal., 1996). Moreover, longitudinal data suggest that negative com-munication indices provide much more discriminative and predic-tive power for marital distress and dissolution than positive be-haviors (Markman, 1991). However, it should be noted that in eachcase the experimental task involved resolving a disagreement, aparadigm that promotes negativity and simultaneously provideslimited opportunities to display supportive behavior. Recent dataemphasize the importance of assessing both conflict and supportbehaviors (Bradbury, Cohan, & Karney, 1998). For example, thesocially supportive behaviors displayed by newlywed spouses asthe partner discussed a personal issue were associated with maritaloutcome 24 months later after controlling for initial levels ofmarital adjustment (Pasch & Bradbury, 1998). By assessing bothconflictual and supportive behaviors, it may be possible to deter-mine whether behavioral problems are unique to conflictual inter-actions or whether they reflect other interpersonal deficits (Brad-bury et al., 1998). In particular, the impact of conflict behaviorsmight be buffered in more supportive marriages but exacerbated in


those where support is low (Bradbury et al., 1998). Bolstering thisview, the strongest cross-sectional predictor of marital disagree-ment frequency was the extent to which one's spouse was per-ceived as supportive (McGonagle & Schilling, 1992).

The Marital Adjustment Test and the DAS, the two most com-monly used marital scales (Locke & Wallace, 1959; Spanier,1976), offer obvious advantages for research on marriage andhealth. They can serve as sensitive and specific measures ofmarital distress (Eddy, Heyman, & Weiss, 1991), and they providenormative data about a sample's marital adjustment. However,their multidimensional nature (both tap several relationship dimen-sions, including satisfaction and disagreement) is problematic, inthat marital quality is not distinguished from related concepts(Eddy et al., 1991). Moreover, conceptualizing marital satisfactionas bidimensional has been criticized, and the need to disaggregatethe assessment of both positive and negative aspects of maritalfunctioning appears to be particularly important; the separability ofpositive and negative components is an important theme in relatedareas of psychological inquiry (Fincham & Beach, 1999). Datafrom the marital interaction studies discussed earlier suggest thathostile or negative behaviors markedly enhance physiologicalchange, particularly among women. Although marital adjustmentdata provide one indirect link, more focused assessments of maritalstrain, conflict, or upset would be useful, particularly in view of thegender differences that may be most obvious when observed in thecontext of the frequency or intensity of marital conflict (Helgeson,1994; Smith et al., 1998).

Relatedly, preliminary evidence supports the value of disaggre-gating positive aspects of marital functioning, particularly in thestudy of health outcomes. For example, specific positive spousalinteractions, but not spousal support, were implicated in diseaseactivity of women with rheumatoid arthritis (Zautra et al., 1998).Without measuring specific positive interactions in this study, theresults may have generated the erroneous conclusion that positivemarital factors were not related to disease activity. In addition, theliterature on marriage and pain outcomes suggests that measuringspecific spousal behaviors along with global marital quality, andassessing the interaction of these two factors, can increase explan-atory power. Levels of global marital satisfaction may moderatepatients' appraisals of specific spousal behaviors, thereby alteringtheir functional significance for health outcomes.

Another recurrent theme in many of the studies reviewed abovewas the importance of including a sufficient number of distressedcouples in study samples; truncation of range may lead to under-estimates of the effect of marital unhappiness. Similarly, routineassessment of marital adjustment, particularly by using scales withknown norms, will enable investigators to assess the extent towhich reductions in range could affect results.

With regard to the two most commonly used approaches tostudying the psychophysiology of marriage (conflict interactionsvs. experimental-impersonal discussions), the former has the ad-vantage of lending greater ecological validity, as couples discusstopics that are of personal relevance and importance to them andthat presumably provoke conflict discussions in their day-to-dayinteractions outside the laboratory (Smith & Gallo, 1999). Exper-imental approaches, although providing less ecological validity,enable greater and more precise control over specific aspects ofmarital interaction (e.g., level of disagreement) that may be im-portant to isolate. Ultimately, the most solid empirical founda-

tion for understanding the psychophysiology of marriage willdevelop from the convergence of multiple methods, includinglaboratory conflict resolution tasks, supportive interactions, andexperimental-impersonal discussions, along with ambulatorystudies that examine marital functioning within the context ofdaily life.

Concerning gender differences, this review of the past decade ofmarriage and health research suggests that future research in thisarena might profitably adopt approaches that are increasinglyprominent in biobehavioral research more broadly: Conduct gen-der comparisons with regard to health outcomes and with regard tobiobehavioral processes that contribute to physical health status(Blanchard, Griebel, & Blanchard, 1995; Legato, 1997). It seemslikely that to have the strongest explanatory power, models ofmarital functioning and health will have to include factors that canaccount for gender differences in biopsychosocial pathways andhealth outcomes. Data from physiological correlates of maritalconflict support this recommendation, and it is strongly suggestedby contemporary psychological models of gender that revealprominent differences in the way relationships affect men's andwomen's self-representations, social interactions, health habits,and risk for depression. Contemporary conceptual approaches togendered traits, self-processes, and marital roles will provide mar-ital researchers with useful starting points for gender-specificassessment, model building, and hypothesis testing.

Finally, considering evidence that marital distress has the po-tential to generate substantial health challenges, can this risk factorbe effectively modified? Although couples therapy reliably in-creases marital satisfaction above that of couples in no-treatmentcontrol groups, some evidence suggests that treatment movescouples from a distressed to a nondistressed range in less than halfthe cases (Christensen & Heavey, 1999). Because significant andlasting endocrine alterations subsequent to a period of conflictoccur even among healthy couples who are predominantly notdistressed (Kiecolt-Glaser, Newton, et al., 1996), many coupleswho participate in marital therapy could continue to experiencelevels of distress that could conceivably contribute to health risk.This seems particularly likely given the chronicity of exposure tomarital distress and conflict, and because improvements in maritalsatisfaction may decay within a few years following termination oftreatment (Christensen & Heavey, 1999). In addition, interventionprograms designed to prevent marital distress have not consistentlybeen interpreted as efficacious (Christensen & Heavey, 1999).Even with an optimistic interpretation of these outcome data, someevidence reveals that married couples with the greatest need ofprevention programs are also the least likely to participate in them(Fincham & Beach, 1999). Nevertheless, precisely because maritaldistress appears to be a potent correlate of health risk factors, thecouples that do benefit substantially from marital therapy mightalso experience notable reductions in health risks; the alcoholliterature has provided some suggestive data in this regard (Sho-ham, Rohrbaugh, Stickle, & Jacob, 1998). Thus, continued effortsat developing efficacious interventions have the potential for siz-able yields in the arena of physical health.

Conclusion

Although both the quality and quantity of social ties have beenrelated to morbidity and mortality (House et al., 1988), the support


provided by certain key relationships is obviously more importantthan others. Indeed, data from national surveys suggest that maritalhappiness contributes far more to global happiness than any othervariable, including satisfaction with work and friendships; how-ever, the relationship is (once again) stronger for wives thanhusbands (Glenn & Weaver, 1981). Although this literature reviewhas emphasized the costs of marital discord, Glenn and Weaverframed the issue somewhat differently and suggested a provocativeand important hypothesis for future studies on marriage and health;they suggested that if women, on average, experience more stressfrom marriage than do men, then the happiness data also suggestthat men derive a lesser benefit from a satisfying relationship—that is, women surpass men in both the stress and satisfactionsgleaned from marriage.

References

Acitelli, L. K., & Young, A. M. (1996). Gender and thought in relation-ships. In G. J. O. Fletcher & J. Fitness (Eds.), Knowledge structures inclose relationships: A social psychological approach (pp. 147-168).Mahwah, NJ: Erlbaum.

Almeida, D. M., & Kessler, R. C. (1998). Everyday stressors and genderdifferences in daily distress. Journal of Personality and Social Psychol-ogy, 75, 670-680.

Alp, M. H., Court, J. H., & Grant, A. K. (1970). Personality pattern andemotional stress in the genesis of gastric ulcer. Gut, 11, 773-777.

American Psychiatric Association. (1987). Diagnostic and statistical man-ual of mental disorders (3rd ed., rev.). Washington, DC: Author.

Appelberg, K., Romanov, K., Heikkila, K., Honkasaol, M. L., & Kosken-vuo, M. (1996). Interpersonal conflict as a predictor of work disability:A follow-up study of 15,348 Finnish employees. Journal of Psychoso-matic Research, 40, 157-167.

Appelberg, K., Romanov, K., Honkasalo, M. L., & Koskenvuo, M. (1993).The use of tranquilizers, hypnotics and analgesics among 18,592 Finnishadults: Associations with recent interpersonal conflicts at work or witha spouse. Journal of Clinical Epidemiology, 46, 1315-1322.

Ashmore, R. D. (1990). Sex, gender, and the individual. In L. A. Pervin(Ed.), Handbook of personality theory and research (pp. 486-526). NewYork: Guilford Press.

Baker, B., Helmers, K., O'Kelly, B., Sakinofsky, I., Abelsohn, A., & Tobe,S. (1999). Marital cohesion and ambulatory blood pressure in earlyhypertension. American Journal of Hypertension, 12, 227-230.

Baker, B., O'Kelly, B., Szalai, J. P., Katie, M., McKessock, D., Ogilvie, R.,Basinski, A., & Tobe, S. W. (1998). Determinants of left ventricularmass in early hypertension. American Journal of Hypertension, I I ,1248-1251.

Barnett, R. C., Davidson, H., & Marshall, N. L. (1991). Physical symptomsand the interplay of work and family roles. Health Psychology, 10,94-101.

Baum, A., Cohen, L., & Hall, M. (1993). Control and intrusive memoriesas possible determinants of chronic stress. Psychosomatic Medicine, 55,274-286.

Baumeister, R. F., & Sommer, K. L. (1997). What do men want? Genderdifferences and two spheres of belongingness: Comment on Cross andMadson (1997). Psychological Bulletin, 122, 38-44.

Beach, S. R. H., Fincham, F. D., & Katz, J. (1998). Marital therapy in thetreatment of depression: Toward a third generation of therapy andresearch. Clinical Psychology Review, 18, 635-661.

Berkman, L. F. (1998). Psychosocial experiences influence functioning:New risks, new outcomes. Psychosomatic Medicine, 60, 256-257.

Berkman, L. F., & Breslow, L. (1983). Health and ways of living: TheAlameda county study. New York: Oxford University Press.

Bird, C. E. (1999). Gender, household labor, and psychological distress:

The impact of the amount and division of housework. Journal of Healthand Social Behavior, 40, 32-45.

Blanchard, D. C., Griebel, G., & Blanchard, R. J. (1995). Gender bias in thepreclinical psychopharmacology of anxiety: Male models for (predom-inately) female disorders. Journal of Psychopharmacology, 9, 79-82.

Bolger, N., Delongis, A., Kessler, R. C., & Schilling, E. A. (1989). Effectsof daily stress on negative mood. Journal of Personality and SocialPsychology, 57, 808-818.

Bolger, N., DeLongis, A., Kessler, R. C., & Wethington, E. (1989). Thecontagion of stress across multiple roles. Journal of Marriage and theFamily, 51, 175-183.

Bradbury, T. N., Beach, S. R. H., Fincham, F. D., & Nelson, G. M. (1996).Attributions and behavior in functional and dysfunctional marriages.Journal of Consulting and Clinical Psychology, 64, 569-576.

Bradbury, T. N., Cohan, C. L., & Karney, B. R. (1998). Optimizinglongitudinal research for understanding and preventing marital dysfunc-tion. In T. N. Bradbury (Ed.), The developmental course of maritaldysfunction (pp. 279-311). New York: Cambridge University Press.

Bradbury, T. N., & Fincham, F. D. (1988). Individual difference variablesin close relationships: A contextual model of marriage as an integrativeframework. Journal of Personality and Social Psychology, 54,113-721.

Bradbury, T. N., & Karney, B. R. (1993). Longitudinal study of maritalinteraction and dysfunction: Review and analysis. Clinical PsychologyReview, 13, 15-27.

Brisson, C., Laflamme, N., Moisan, J., Milot, A., Masse, B., & Vezina, M.(1999). Impact of family responsibilities and job strain on ambulatoryblood pressure among white-collar women. Psychosomatic Medi-cine, 61, 205-213.

Broadwell, S. D., & Light, K. C. (1999). Family support and cardiovascularresponses in married couples during conflict and other interactions.International Journal of Behavioral Medicine, 6, 40-63.

Brown, P. C., & Smith, T. W. (1992). Social influence, marriage, and theheart: Cardiovascular consequences of interpersonal control in husbandsand wives. Health Psychology, 11, 88-96.

Brown, P. C., Smith, T. W., & Benjamin, L. S. (1998). Perceptions ofspouse dominance predict blood pressure reactivity during marital in-teractions. Annals of Behavioral Medicine, 20, 286-293.

Brummett, B. H., Barefoot, J. C., Feaganes, J. R., Yen, S., Bosworth, H. B.,Williams, R. B., & Siegler, I. C. (2000). Hostility in marital dyads:Associations with depressive symptoms. Journal of Behavioral Medi-cine, 23, 95-105.

Burman, B., & Margolin, G. (1992). Analysis of the association betweenmarital relationships and health problems: An interactional perspective.Psychological Bulletin, 112, 39-63.

Burnett, R. (1987). Reflections in personal relationships. In R. Burnett, P.McGhee, & D. Clarke (Eds.), Accounting for relationships: Explanation,representation, consciousness (pp. 102-110). New York: Methuen.

Burns, E. A., & Goodwin, J. S. (1990). Immunology and infectious disease.In C. K. Cassel, D. E. Risenberg, L. B. Sorensen, & J. R. Walsh (Eds.),Geriatric medicine (pp. 312-329). New York: Springer-Verlag.

Burns, J. W., Johnson, B. J., Mahoney, N., Devine, J., & Pawl, R. (1996).Anger management style, hostility and spouse responses: Gender differ-ences in predictors of adjustment among chronic pain patients. Pain, 64,445-453.

Carels, R. A., Sherwood, A., & Blumenthal, J. A. (1998). Psychosocialinfluences on blood pressure during daily life. International Journal ofPsychophysiology, 28, 117-129.

Chandra, V., Szklo, M., Goldberg, R., & Tonascia, J. (1983). The impactof marital status on survival after an acute myocardial infarction: Apopulation-based study. American Journal of Epidemiology, 117, 320-325.

Christensen, A. (1987). Detection of conflict patterns in couples. In K.Hahlweg & M. Goldstein (Eds.), Understanding major mental disorder(pp. 250-265). New York: Family Process.


Christensen, A., & Heavey, C. L. (1999). Interventions for couples. AnnualReview of Psychology, 50, 165-190.

Christensen, A., & Margolin, G. (1988). Conflict and alliance in distressedand nondistressed families. In R. A. Hinde & J. Stevenson-Hinde (Eds.),Relationships within families: Mutual influences (pp. 263-282). Oxford,England: Clarendon.

Cohen, S., Frank, E., Doyle, W. J., Skoner, D. P., Rabin, B. S., &Gwaltney, J. M. (1998). Types of stressors that increase susceptibility tothe common cold in healthy adults. Health Psychology, 17, 214-223.

Cohen, S., Schwartz, J. E., Bromet, E. J., & Parkinson, D. K. (1991).Mental health, stress, and poor health behaviors in two communitysamples. Preventive Medicine, 20, 306-315.

Coughlin, P. C. (1990). Premenstrual syndrome: How marital satisfactionand role choice affect symptom severity. Social Work, 35, 351-355.

Coyne, J. C., & DeLongis, A. (1986). Going beyond social support: Therole of social relationships in adaptation. Journal of Consulting andClinical Psychology, 54, 454-460.

Cross, S. E., & Madson, L. (1997a). Elaboration of models of the self:Reply to Baumeister and Sommer (1997) and Martin and Ruble (1997).Psychological Bulletin, 122, 51-55.

Cross, S. E., & Madson, L. (1997b). Models of the self: Self-construals andgender. Psychological Bulletin, 122, 5-37.

DeRijk, R., Michelson, D., Karp, B., Petrides, J., Galliven, E., Deuster, P.,Paciotti, G., Gold, P. W., & Sternberg, E. M. (1997). Exercise andcircadian rhythm-induced variations in plasma cortisol differentiallyregulate Interleukin-lB (IL-1/3), IL-6, and Tumor Necrosis Factor-a(TNF-a) production in humans: High sensitivity of TNF-a and resis-tance of IL-6. Journal of Clinical Endocrinology and Metabolism, 82,2182-2192.

Dhabhar, F. S., & McEwen, B. S. (1997). Acute stress enhances whilechronic stresses suppresses cell-mediated immunity in vivo: A potentialrole for leukocyte trafficking. Brain, Behavior and Immunity, 11, 286-306.

Earle, T. L., Linden, W., & Weinberg, J. (1999). Differential effects ofharassment on cardiovascular and salivary cortisol stress reactivity andrecovery in women and men. Journal of Psychosomatic Research, 46,125-141.

Eddy, J. M., Heyman, R. E., & Weiss, R. L. (1991). An empiricalevaluation of the Dyadic Adjustment Scale: Exploring the differencesbetween marital "satisfaction" and "adjustment." Behavioral Assess-ment, 13, 199-220.

Ewart, C. K. (1993). Marital interaction—The context for psychosomaticresearch. Psychosomatic Medicine, 55, 410-412.

Ewart, C. K., Taylor, C. B., Kraemer, H. C., & Agras, W. S. (1984).Reducing blood pressure reactivity during interpersonal conflict: Effectsof marital communication training. Behavior Therapy, 15, 473-484.

Ewart, C. K., Taylor, C. B., Kraemer, H. C., & Agras, W. S. (1991). Highblood pressure and marital discord: Not being nasty matters more thanbeing nice. Health Psychology, 10, 155-163.

Fehm-Wolfsdorf, G., Groth, T., Kaiser, A., & Hahlweg, K. (1999). Cortisolresponses to marital conflict depend on marital interaction quality.International Journal of Behavioral Medicine, 6, 207—227.

Fincham, F. D., & Beach, S. R. (1999). Conflict in marriage: Implicationsfor working with couples. Annual Review of Psychology, 50, 41-11.

Fincham, F. D., Beach, S. R. H., Harold, G. T., & Osborne, L. N. (1997).Marital satisfaction and depression: Different causal relationships formen and women? Psychological Science, 8, 351-357.

Fisher, L., Nakell, L. C., Terry, H. E., & Ransom, D. C. (1992). TheCalifornia Family Health Project: III. Family emotion management andadult health. Family Process, 31, 269-287.

Flor, H., Breitenstein, C., Birbaumer, N., & Furst, M. (1995). A psycho-physiological analysis of spouse solicitousness towards pain behaviors,spouse interaction, and pain perception. Behavior Therapy, 26, 255-272.

Floyd, F. J., & Markman, H. J. (1983). Observational biases in spousal

observations: Toward a cognitive behavioral model of marriage. Journalof Consulting and Clinical Psychology, 51, 450-457.

Frankenhaeuser, M. (1986). A psychobiological framework for research onhuman stress and coping. In M. H. Appley & R. Trumbull (Eds.),Dynamics of stress: Physiological, psychological, and social perspec-tives (pp. 101-116). New York: Plenum.

Frankenhauser, M., Lundberg, U., Fredrikson, M., Melin, B., & Tuomisto,M. (1989). Stress on and off the job as related to sex and occupationalstatus in white collar workers. Journal of Organizational Behavior, 10,321-346.

Prankish, C. J., & Linden, W. (1996). Spouse-pair risk factors and cardio-vascular reactivity. Journal of Psychosomatic Research, 40, 37-51.

Fredrikson, M., Tuomisto, M., & Bergman-Losman, B. (1991). Neuroen-docrine and cardiovascular stress reactivity in middle-age normotensiveadults with parental history of cardiovascular disease. Psychophysiol-ogy, 28, 656-664.

Gabriel, S., & Gardner, W. L. (1999). Are there "his" and "hers" types ofinterdependence? The implications of gender differences in collectiveversus relational independence for affect, behavior, and cognition. Jour-nal of Personality and Social Psychology, 77, 642-655.

Gaelick, L., Bodenhausen, G. V., & Wyer, R. S. J. (1985). Emotionalcommunication in close relationships. Journal of Personality and SocialPsychology, 49, 1246-1265.

Ganong, L. H., & Coleman, M. (1991). Remarriage and health. Researchin Nursing and Health, 14, 205-211.

Geiss, A., Varadi, E., Steinbach, K., Bauer, H. W., & Anton, F. (1997).Psychoneuroimmunological correlates of persisting sciatic pain in pa-tients who underwent discectomy. Neuroscience Letters, 237, 65-68.

Gillies, M., & Skyring, A. (1968). Gastric ulcer, duodenal ulcer and gastriccarcinoma: A case-control study of certain social and environmentalfactors. Medical Journal of Australia, 2, 1132-1136.

Glaser, R., & Kiecolt-Glaser, J. K. (Eds.). (1994). Handbook of humanstress and immunity. San Diego, CA: Academic Press.

Glaser, R., Kiecolt-Glaser, J. K., Bonneau, R., Malarkey, W., & Hughes, J.(1992). Stress-induced modulation of the immune response to recombi-nant Hepatitis B vaccine. Psychosomatic Medicine, 54, 22-29.

Glaser, R., Kiecolt-Glaser, J. K., Marucha, P. T., MacCallum, R. C.,Laskowski, B. F., & Malarkey, W. B. (1999). Stress-related changes inproinflammatory cytokine production in wounds. Archives of GeneralPsychiatry, 56, 450-456.

Glass, J., & Fujimoto, T. (1994). Housework, paid work, and depressionamong husbands and wives. Journal of Health and Social Behavior, 35,179-191.

Glassman, A. H., & Shapiro, P. A. (1998). Depression and the course ofcoronary artery disease. American Journal of Psychiatry, 155, 4-11.

Glenn, N. D., & Weaver, C. N. (1981). The contribution of maritalhappiness to global happiness. Journal of Marriage and the Family, 43,161-168.

Goldstein, I. B., Shapiro, D., Chicz-DeMet, A., & Guthrie, D. (1999).Ambulatory blood pressure, heart rate, and neuroendocrine responses inwomen nurses during work and off work days. Psychosomatic Medi-cine, 61, 387-396.

Goodwin, J. S., Hunt, W. C., Key, C. R., & Samet, J. M. (1987). The effectof marital status on stage, treatment, and survival of cancer patients.Journal of the American Medical Association, 34, 20-26.

Goodwin, S. (1997). The marital relationship and health in women withchronic fatigue and immune dysfunction syndrome: Views of wives andhusbands. Nursing Research, 46, 138-146.

Gordon, H. S., & Rosenthal, G. E. (1995). Impact of marital status onoutcomes in hospitalized patients. Archives of Internal Medicine, 155,2465-2471.

Gottman, J. M., & Levenson, R. W. (1992). Marital processes predictive oflater dissolution: Behavior, physiology, and health. Journal of Person-ality and Social Psychology, 63, 221-233.


Greene, S. M., & Griffin, W. A. (1998). Symptom study in context: Effectsof marital quality on signs of Parkinson's disease during patient-spouseinteraction. Psychiatry, 61, 35-45.

Greisen, J., Hokland, M., Grfote, T., Hansen, P. O., Jensen, T. S., Vilstrup,H., &-Tonnesen, E. (1999). Acute pain induces an instant increase innatural killer cell cytotoxicity in humans and this response is abolishedby local anaesthesia. British Journal of Anaesthesia, 83, 235-240.

Hafner, R. J., Rogers, J., & Watts, J. M. (1990). Psychological status beforeand after gastric restriction as predictors of weight loss in the morbidlyobese. Journal of Psychosomatic Research, 34, 295-302.

Hagestad, G. O., & Smyer, M. A. (1982). Dissolving long term relation-ships: Patterns of divorcing in middle age. In S. Duck (Ed.), Personalrelationships: Vol. 4. Dissolving relationships (pp. 115-188). NewYork: Academic Press.

Harvey, J. H., Wells, G. L., & Alvarez, M. D. (1978). Attribution in thecontext of conflict and separation in close relationships. In J. H. Harvey,W. Ickes, & R. F. Kidd (Eds.), New directions in attribution research(Vol. 2, pp. 235-260). Hillsdale, NJ: Erlbaum.

Heavey, C. L., Layne, C., & Christensen, A. A. (1993). Gender and conflictstructure in marital interaction: A replication and extension. Journal ofConsulting and Clinical Psychology, 61, 16-27.

Helgeson, V. S. (1991). The effects of masculinity and social support onrecovery from myocardial infarction. Psychosomatic Medicine, 53, 621-633.

Helgeson, V. (1993). Implications of agency and communion for patientand spouse adjustment to a first coronary event. Journal of Personalityand Social Psychology, 64, 807-816.

Helgeson, V. S. (1994). Relation of agency and communion to well-being:Evidence and potential explanations. Psychological Bulletin, 116, 416-428.

Helgeson, V., & Fritz, H. (1996). Implications of communion and unmit-igated communion for adolescent adjustment to Type I diabetes. Wom-en's Health: Research on Gender, Behavior, and Policy, 2, 169-194.

Herrmann, C., Brand-Driehorst, S., Kaminsky, B., Leibing, E., Staats, H.,& Ruger, U. (1998). Diagnostic groups and depressed mood as predic-tors of 22-month mortality in medical inpatients. Psychosomatic Medi-cine, 60, 570-577.

Hibbard, J. H., & Pope, C. R. (1993). The quality of social roles aspredictors of morbidity and mortality. Social Science and Medicine, 36,217-225.

Hobson, D., Curry, R. L., & Beare, A. S. (1972). Hemagglutination-inhibiting antibody litres as a measure of protection against influenza inman. In F. T. Perkins & R. H. Reganey (Eds.), International symposiumon influenza vaccines for men and horses (pp. 164-168). New York:Karger.

Hooley, J. M., & Teasdale, J. D. (1989). Predictors of relapse in unipolardepressives: Expressed emotion, marital distress, and perceived criti-cism. Journal of Abnormal Psychology, 98, 229-235.

Horwitz, A. V., & White, H. R. (1991). Becoming married, depression, andalcohol problems among young adults. Journal of Health and SocialBehavior, 32, 221-237.

House, J. S., Landis, K. R., & Umberson, D. (1988, July 29). Socialrelationships and health. Science, 241, 540-545.

Houston, B. K., & Kelly, K. E. (1989). Hostility in employed women:Relation to work and marital experiences, social support, stress, andanger expression. Personality and Social Psychology Bulletin, 15, 175-182.

Hiibner, G., Brauchle, M., Smola, H., Madlener, M., Fassler, R., & Werner,S. (1996). Differential regulation of pro-inflammatory cytokines duringwound healing in normal and glucocorticoid-treated mice. Cytokine, 8,548-556.

Idler, E. L., & Benyamini, Y. (1997). Self-rated health and mortality: Areview of twenty-seven community studies. Journal of Health andSocial Behavior, 38, 21-37.

Jacob, T., & Leonard, K. (1988). Alcoholic—spouse interaction as a func-tion of alcoholism subtype and alcohol consumption interaction. Journalof Abnormal Psychology, 97, 231-237.

Jacob, T., & Leonard, K. (1992). Sequential analysis of marital interactionsinvolving alcoholic, depressed and nondistressed men. Journal of Ab-normal Psychology, 101, 547-565.

Jacobson, N. S., Gottman, J. M., Waltz, J., Rushe, R., Bobcock, J., &Holtzworth-Monroe, A. (1994). Affect, verbal content, and psychophys-iology in the arguments of couples with a violent husband. Journal ofConsulting and Clinical Psychology, 62, 982-988.

James, G. D., Gates, E. M., Pickering, T. G., & Laragh, J. H. (1989). Parityand perceived job stress elevate blood pressure in young normotensiveworking women. American Journal of Hypertension, 2, 637-689.

Jenkins, C. D., Kraeger, B. E., Rose, R. M., & Hurst, M. W. (1980). Useof a monthly health review to ascertain illness and injuries. AmericanJournal of Public Health, 70, 82-84.

Jones, D. A., Rollman, G. B., & Brooke, R. I. (1997). The cortisol responseto psychological stress in temporomandibular dysfunction. Pain, 72,171-182.

Keefe, F. J., Caldwell, D. S., Baucom, D., Salley, A., Robinson, E.,Timmons, K., Beaupre, P., Weisberg, J., & Helms, M. (1996). Spouse-assisted coping skills training in the management of osteoarthritic kneepain. Arthritis Care and Research, 9, 279-291.

Keefe, F. J., Caldwell, D. S., Baucom, D., Salley, A., Robinson, E.,Timmons, K., Beaupre, P., Weisberg, J., & Helms, M. (1999). Spouse-assisted coping skills training in the management of knee pain inosteoarthritis: Long-term follow-up results. Arthritis Care and Re-search, 12, 101-111.

Kelly, A. B., & Halford, W. K. (1995). The generalisation of cognitivebehavioural marital therapy in behavioural, cognitive and physiologicaldomains. Behavioural and Cognitive Psychotherapy, 23, 381-398.

Kerns, R. D., Haythornthwaite, J., Southwick, S., & Ciller, E. L. (1990).The role of marital interaction in chronic pain and depressive symptomseverity. Journal of Psychosomatic Research, 34, 401-408.

Kiecolt-Glaser, J. K. (1999). Stress, personal relationships, and immunefunction: Health implications. Brain, Behavior and Immunity, 13, 61-72.

Kiecolt-Glaser, J. K., Cacioppo, J. T., Malarkey, W. B., & Glaser, R.(1992). Acute psychological stressors and short-term immune changes:What, why, for whom, and to what extent? Psychosomatic Medicine, 54,680-685.

Kiecolt-Glaser, J. K., Fisher, L., Ogrocki, P., Stout, J. C., Speicher, C. E.,& Glaser, R. (1987). Marital quality, marital disruption, and immunefunction. Psychosomatic Medicine, 49, 31-34.

Kiecolt-Glaser, J. K., & Glaser, R. (1988). Methodological issues inbehavioral immunology research with humans. Brain, Behavior, andImmunity, 2, 67-78.

Kiecolt-Glaser, J. K., Glaser, R., Cacioppo, J. T., MacCallum, R. C.,Snydersmith, M., Kim, C., & Malarkey, W. B. (1997). Marital conflictin older adults: Endocrinological and immunological correlates. Psycho-somatic Medicine, 59, 339-349.

Kiecolt-Glaser, J. K., Glaser, R., Cacioppo, J. T., & Malarkey, W. B.(1998). Marital stress: Immunologic, neuroendocrine, and autonomiccorrelates. Annals of the New York Academy of Sciences, 840, 649-655.

Kiecolt-Glaser, J. K., Glaser, R., Gravenstein, S., Malarkey, W. B., &Sheridan, J. (1996). Chronic stress alters the immune response to influ-enza virus vaccine in older adults. Proceedings of the National Academyof Sciences, USA, 93, 3043-3047.

Kiecolt-Glaser, J. K., Kennedy, S., Malkoff, S., Fisher, L., Speicher, C. E.,& Glaser, R. (1988). Marital discord and immunity in males. Psycho-somatic Medicine, 50, 213-229.

Kiecolt-Glaser, J. K., Malarkey, W. B., Chee, M., Newton, T., Cacioppo,J. T., Mao, H., & Glaser, R. (1993). Negative behavior during maritalconflict is associated with immunological down-regulation. Psychoso-matic Medicine, 55, 395-409.


Kiecolt-Glaser, J. K., Marucha, P. T., Malarkey, W. B., Mercado, A. M., &Glaser, R. (1995). Slowing of wound healing by psychological stress.Lancet, 346, 1194-1196.

Kiecolt-Glaser, J. K., Newton, T., Cacioppo, J. T., MacCallum, R. C.,Glaser, R., & Malarkey, W. B. (1996). Marital conflict and endocrinefunction: Are men really more physiologically affected than women?Journal of Consulting and Clinical Psychology, 64, 324-332.

Kiecolt-Glaser, J. K., Page, G. G., Marucha, P. T., MacCallum, R. C., &Glaser, R. (1998). Psychological influences on surgical recovery: Per-spectives from psychoneuroimmunology. American Psychologist, 53,1209-1218.

Kirschbaum, C., Wust, S., & Hellhammer, D. (1992). Consistent sexdifferences in cortisol responses to psychological stress. PsychosomaticMedicine, 54, 648-657.

Kitson, C. G. (1982). Attachment to the spouse in divorce: A scale and itsapplication. Journal of Marriage and the Family, 44, 379-393.

Kuhn, C. M. (1989). Adrenocortical and gonadal steroids in behavioralcardiovascular medicine. In N. Schneiderman, S. M. Weiss, & P. G.Kaufman (Eds.), Handbook of research methods in cardiovascular be-havioral medicine (pp. 185-204). New York: Plenum.

Kulik, J. A., & Mahler, H. I. (1989). Social support and recovery fromsurgery. Health Psychology, 8, 221-238.

Larson, R. W., & Almeida, D. M. (1999). Emotional transmission in thedaily lives of families: A new paradigm for studying family process.Journal of Marriage and the Family, 61, 5-20.

Legato, M. J. (1997). Gender-specific physiology: How real is it? Howimportant is it? International Journal of Fertility, 42, 19-29.

Lentjes, E. G., Griep, E. N., Boersma, J. W., Romijn, F. P., & de Kloet,E. R. (1997). Glucocorticoid receptors, fibromyalgia and low back pain.Psychoneuroendocrinology, 22, 603-614.

LeResche, L. (1995). Gender differences in pain: Epidemiologic perspec-tives. Pain Forum, 4, 228-230.

Levenson, R. W., Carstensen, L. L., & Gottman, J. M. (1993). Long-termmarriage: Age, gender, and satisfaction. Psychology and Aging, 2, 301-313.

Levenson, R. W., Carstensen, L. L., & Gottman, J. M. (1994). Theinfluence of age and gender on affect, physiology, and their interrela-tions: A study of long-term marriages. Journal of Personality and SocialPsychology, 45, 587-597.

Levenstein, S., Ackerman, S., Kiecolt-Glaser, J. K., & Dubois, A. (1999).Stress and peptic ulcer disease. Journal of the American Medical Asso-ciation, 281, 10-11.

Levenstein, S., Kaplan, G. A., & Smith, M. (1995). Sociodemographiccharacteristics, life stressors, and peptic ulcer. A prospective study.Journal of Clinical Gastroenterology, 21, 185-192.

Lewis, M. A., Rook, K. S., & Schwarzer, R. (1994). Social support, socialcontrol, and health among the elderly. In G. N. Penny, P. Bennett, & M.Herbert (Eds.), Health psychology: A lifespan perspective (pp. 191-211). Philadelphia: Harwood Academic.

Liebeskind, J. C. (1991). Pain can kill. Pain, 44, 3-4.Litwak, E., & Messeri, P. (1989). Organizational theory, social supports,

and mortality rates: A theoretical convergence. American SociologicalReview, 54, 49-66.

Locke, H. J., & Wallace, K. M. (1959). Short marital adjustment andprediction tests: Their reliability and validity. Marriage and FamilyLiving, 21, 251-255.

Lousberg, R., Schmidt, A. J., & Groenman, N. H. (1992). The relationshipbetween spouse solicitousness and pain behavior: Searching for moreexperimental evidence. Pain, 51, 75-79.

Luecken, L. J., Suarez, E. C., Kuhn, C. M., Barefoot, J. C., Blumenthal,J. A., Siegler, I. C., & Williams, R. B. (1997). Stress in employedwomen: Impact of marital status and children at home on neurohormoneoutput and home strain. Psychosomatic Medicine, 59, 352-359.

Malarkey, W., Kiecolt-Glaser, J. K., Pearl, D., & Glaser, R. (1994). Hostile

behavior during marital conflict alters pituitary and adrenal hormones.Psychosomatic Medicine, 56, 41-51.

Marcenes, W., & Sheiham, A. (1996). The relationship between maritalquality and oral health status. Psychology and Health, 11, 357-369.

Margolin, G., Christensen, A., & John, R. S. (1996). The continuance andspillover of everyday tensions in distressed and nondistressed families.Journal of Family Psychology, 10, 304-321.

Markman, H. J. (1991). Constructive marital conflict is NOT an oxymoron.Behavioral Assessment, 13, 83-96.

Marucha, P. T., Kiecolt-Glaser, J. K., & Favagehi, M. (1998). Mucosalwound healing is impaired by examination stress. Psychosomatic Med-icine, 60, 362-365.

Mayne, T. J., O'Leary, A., McCrady, B., Contrada, R., & Labouvie, E.(1997). The differential effects of acute marital distress on emotional,physiological and immune functions in maritally distressed men andwomen. Psychology and Health, 12, 277-288.

McGonagle, K. A., & Schilling, E. A. (1992). The frequency and deter-minants of marital disagreements in a community sample. Journal ofSocial and Personal Relationships, 9, 507-524.

Meana, M., Binick, L, Khalife, S., & Cohen, D. (1998). Affect and maritaladjustment in women's rating of dyspareunic pain. Canadian Journal ofPsychiatry, 43, 381-385.

Mechanic, D. (1980). The experience and reporting of common physicalcomplaints. Journal of Health and Social Behavior, 21, 146-155.

Medalie, J. H., Stange, K. C., Zyanski, S. J., & Goldbourt, U. (1992). Theimportance of biopsychosocial factors in the development of duodenalulcer in a cohort of middle-aged men. American Journal of Epidemiol-ogy, 136, 1280-1287.

Miller, G. E., Dopp, J. M., Myers, H. F., Felten, S. Y., & Fahey, J. L.(1999). Psychosocial predictors of natural killer cell mobilization duringmarital conflict. Health Psychology, 18, 262-271.

Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M. L., & Hallet, A. J.(1996). A meta-analytic review of research on hostility and physicalhealth. Psychological Bulletin, 119, 322-348.

Morell, M. A., & Apple, R. F. (1990). Affect expression, marital satisfac-tion, and stress reactivity among premenopausal women during a con-flictual marital discussion. Psychology of Women Quarterly, 14, 387-402.

Newton, T. L., & Kiecolt-Glaser, J. K. (1995). Hostility and erosion ofmarital quality during early marriage. Journal of Behavioral Medi-cine, 18, 601-619.

Newton, T. L., Kiecolt-Glaser, J. K., Glaser, R., & Malarkey, W. B. (1995).Conflict and withdrawal during marital interaction: The roles of hostilityand defensiveness. Personality and Social Psychology Bulletin, 21,512-524.

Nolen-Hoeksema, S., & Girgus, J. S. (1994). The emergence of genderdifferences in depression during adolescence. Psychological Bulletin,115, 424-443.

Nolen-Hoeksema, S., Larson, J., & Grayson, C. (1999). Explaining thegender difference in depressive symptoms. Journal of Personality andSocial Psychology, 77, 1061-1072.

Noller, P., & Fitzpatrick, M. A. (1990). Marital communication in theeighties. Journal of Marriage and the Family, 52, 832-843.

Notarius, C., Benson, S., Sloane, D., Vanzetti, N., & Hornyak, L. (1989).Exploring the interface between perception and behavior: An analysis ofmarital interaction in distressed and nondistressed couples. BehavioralAssessment, 11, 39-64.

Notarius, C. I., & Johnson, J. S. (1982). Emotional expression in husbandsand wives. Journal of Marriage and the Family, 44, 483-489.

O'Farrell, T., Hooley, J., Fals-Stewart, W., & Cutter, H. S. G. (1998).Expressed emotion and relapse in alcoholic patients. Journal of Con-sulting and Clinical Psychology, 66, 744-752.

O'Leary, K. D., Christian, J. L., & Mendell, N. R. (1994). A closer look at


the link between marital discord and depressive symptomatology. Jour-nal of Social and Clinical Psychology, 13, 33-41.

Orts, K., Sheridan, J. F., Robinson-Whelen, S., Glaser, R., Malarkey,W. B., & Kiecolt-Glaser, J. K. (1995). The reliability and validity of astructured interview for the assessment of infectious illness. Journal ofBehavioral Medicine, 18, 517-530.

Page, G. G.. McDonald, J. S., & Ben-Eliyahu, S. (1998). Pre-operativeversus postoperative administration of morphine: Impact on the neu-roendocrine, behavioral, and metastatic-enhancing effects of surgery.British Journal of Anaesthesia, 81, 216-223.

Pasch, L. A., & Bradbury, T. N. (1998). Social support, conflict, and thedevelopment of marital dysfunction. Journal of Consulting and ClinicalPsychology, 66, 219-230.

Patriarca, P. A. (1994). A randomized controlled trial of influenza vaccinein the elderly. Journal of the American Medical Association, 272,1700-1701.

Paulsen, J. S., & Altmaier, E. M. (1995). The effects of perceived versusenacted social support on the discriminative cue function of spouses forpain behaviors. Pain, 60, 103-110.

Pennebaker, J. W., & Roberts, T. A. (1992). Toward a his and hers theoryof emotion: Gender differences in visceral perception. Journal of Socialand Clinical Psychology, 11, 199-212.

Penninx, B. W. J. H., Guralnik, J. M., Ferrucci, L., Simonsick, E. M., Deeg,D. J. H., & Wallace, R. B. (1998). Depressive symptoms and physicaldecline in community-dwelling older persons. Journal of the AmericanMedical Association, 279, 1720-1726.

Penninx, B. W. J. H., Guralnik, J. M., Pahor, M., Ferrucci, L., Cerhan,J. R., Wallace, R. B., & Havlik, R. J. (1998). Chronically depressedmood and cancer risk in older persons. Journal of the National CancerInstitute, 90, 1888-1893.

Pezzone, M. A., Dohanics, J., & Rabin, B. S. (1994). Effects of footshockstress upon spleen and peripheral blood lymphocyte mitogenic responsesin rats with lesions of the paraventricular nucleus. Journal of Neuroim-munology, 53, 39-46.

Phillipson. C. (1997). Social relationships in later life: A review of theresearch literature. International Journal of Geriatric Psychiatry, 12,505-512.

Prigerson, H. G., Maciejewski, P. K., & Rosenheck, R. A. (1999). Theeffects of marital dissolution and marital quality on health and healthservices use among women. Medical Care, 37, 858-873.

Ren, X. S. (1997). Marital status and quality of relationships: The impacton health perception. Social Science and Medicine, 44, 241-249.

Repetti, R. L. (1989). Effects of daily workload on subsequent behaviorduring marital interaction: The roles of social withdrawal and spousesupport. Journal of Personality and Social Psychology, 57, 651-659.

Roberts, L. J., & Krokoff, L. J. (1990). A time-series analysis of with-drawal, hostility, and displeasure in satisfied and dissatisfied marriages.Journal of Marriage and the Family, 52, 95-105.

Romano, J. M., Turner, J. A., Friedman, L. S., Bulcroft, R. A., Jensen,M. P., Hops, H., & Wright, S. F. (1995). Chronic pain patient-spousebehavioral interactions predict patient disability. Pain, 63, 353-360.

Romano, J. M., Turner. J. A., & Jensen, M. P. (1997). The familyenvironment in chronic pain patients: Comparison to controls and rela-tionship to patient functioning. Journal of Clinical Psychology in Med-ical Settings, 4, 383-395.

Rook, K. S. (1998). Investigating the positive and negative sides ofpersonal relationships: Through a lens darkly. In B. H. Spitzberg &W. R. Capach (Eds.), The dark side of close relationships (pp. 369-393).Mahwah, NJ: Erlbaum.

Ross. C. E., Mirowsky, J., & Goldsteen, K. (1990). The impact of thefamily on health: The decade in review. Journal of Marriage and theFamily, 52, 1059-1078.

Ross. M., & Holmberg, D. (1990). Recounting the past: Gender differencesin the recall of events in the history of a close relationship. In J. M. Olson

& M. P. Zanna (Eds.), Self-influence processes (pp. 135-152). Hillsdale,NJ: Erlbaum.

Roth-Roemer, S., & Kurpius, S. E. R. (1996). Beyond marital status: Anexamination of marital quality and well-being among women withrheumatoid arthritis. Women's Health: Research on Gender, Behavior,and Policy, 2, 195-205.

Saarijarvi, S., Rytokoski, U., & Alanen, E. (1991). A controlled study ofcouple therapy in chronic low back pain patients: No improvement ofdisability. Journal of Psychosomatic Research, 35, 671-677.

Saarijarvi, S., Rytokoski, U., & Karppi, S. L. (1990). Marital satisfactionand distress in chronic low-back pain patients and their spouses. ClinicalJournal of Pain, 6, 148-152.

Santos, J., Saperas, E., Nogueiras, C., Mourelle, M., Antolin, M., Cadahia,A., & Malagelada, J. R. (1998). Release of mast cell mediators into thejejunum by cold pain stress in humans. Gastroenterology, 114, 640-648.

Scherwitz, L., & Rugulies, R. (1992). Life-style and hostility. In H. S.Friedman (Ed.), Hostility, coping, and health (pp. 77-98). WashingtonDC: American Psychological Association.

Schleifer, S. J., Keller, S. E., Bond, R. N., Cohen, J., & Stein, M. (1989).Depression and immunity: Role of age, sex, and severity. Archives ofGeneral Psychiatry, 46, 81-87.

Schmaling, K. B., Wamboldt, F., Telford, L., Newman, K. B., Hops, H., &Eddy, J. M. (1996). Interaction of asthmatics and their spouses: Apreliminary study of individual differences. Journal of Clinical Psychol-ogy in Medical Settings, 3, 211-218.

Schwartz, L., Slater, M. A., & Birchler, G. R. (1994). Interpersonal stressand pain behaviors in patients with chronic pain. Journal of Consultingand Clinical Psychology, 62, 861-864.

Schwartz, L., Slater, M. A., & Birchler, G. R. (1996). The role of painbehaviors in the modulation of marital conflict in chronic pain couples.Pain, 65, 227-233.

Shoham, V., Rohrbaugh, M. J., Stickle, T. R., & Jacob, T. (1998).Demand-withdraw couple interaction moderates retention in cognitive-behavioral versus family-systems treatments for alcoholism. Journal ofFamily Psychology, 12, 557-577.

Simonsick, E. M., Wallace, R. B., Blazer, D. G., & Berkman, L. F. (1995).Depressive symptomatology and hypertension-associated morbidity andmortality in older adults. Psychosomatic Medicine, 57, 427-435.

Smith, T. W. (1992). Hostility and health: Current status of a psychoso-matic hypothesis. Health Psychology, 11, 139-150.

Smith, T. W., & Brown, P. C. (1991). Cynical hostility, attempts to exertsocial control, and cardiovascular reactivity in married couples. Journalof Behavioral Medicine, 14, 581-592.

Smith, T. W., & Gallo, L. C. (1999). Hostility and cardiovascular reactivityduring marital interaction. Psychosomatic Medicine, 61, 436-453.

Smith, T. W., Gallo, L. C., Goble, L., Ngu, L. Q., & Stark, K. A. (1998).Agency, communion, and cardiovascular reactivity during marital inter-action. Health Psychology, 17, 537-545.

Smith, T. W., Pope, M. K., Sanders, J. D., Allred, K. D., & O'Keefe, J. L.(1988). Cynical hostility at home and work: Psychosocial vulnerabilityacross domains. Journal of Research in Personality, 22, 525-548.

Smith, T. W., Sanders, J. D., & Alexander, J. F. (1990). What does theCook and Medley Hostility scale measure? Affect, behavior, and attri-butions in the marital context. Journal of Personality and Social Psy-chology, 58, 699-708.

Spanier, G. B. (1976). Measuring dyadic adjustment: New scales forassessing the quality of marriage and similar dyads. Journal of Marriageand the Family, 38, 15-28.

Stansfeld, S. A., Bosnia, H., Hemingway, H., & Marmot, M. G. (1998).Psychosocial work characteristics and social support as predictors ofSF-36 health functioning: The Whitehall II Study. Psychosomatic Med-icine, 60, 247-255.

Steptoe, A., Lundwall, K., & Cropley, M. (2000). Gender, family structure


and cardiovascular activity during the working day and evening. SocialScience and Medicine, 50, 531-5?9.

Stoney, M. C., Davis, M. C., & Matthews, K. A. (1987). Sex differences inphysiological responses to stress and in coronary heart disease: A causallink? Psychophysiology, 24, 127-131.

Sullivan, M., Katon, W., Russo, J., Dobie, R., & Sakai, C. (1994). Copingand marital support as correlates of tinnitus disability. General HospitalPsychiatry, 16, 259-266.

Suls, J., & Wan, C. K. (1993). The relationship between trait hostility andcardiovascular reactivity: A quantitative review and analysis. Psycho-physiology, 30, 615-626.

Thomas, S. P. (1995). Psychosocial correlates of women's health in middleadulthood. Issues in Mental Health Nursing, 16, 285-314.

Thomsen, D. G., & Gilbert, D. G. (1998). Factors characterizing maritalconflict states and traits: Physiological, affective, behavioral and neu-rotic variable contributions to marital conflict and satisfaction. Person-ality and Individual Differences, 25, 833-855.

Trevino, D. B., Young, E. H., Groff, J., & Jono, R. T. (1990). Theassociation between marital adjustment and compliance with antihyper-tension regimens. Journal of the American Board of Family Practice, 3,17-25.

Turk, D. C., Kerns, R. D., & Rosenberg, R. (1992). Effects of maritalinteraction on chronic pain and disability: Examining the down side ofsocial support. Rehabilitation Psychology, 37, 259-274.

Turk, D. C., & Melzack, R. (1992). The measurement of pain and theassessment of people experiencing pain. In D. C. Turk & R. Melzack(Eds.), Handbook of pain assessment (pp. 3-12). New York: GuilfordPress.

Tuschen-Caffier, B., Florin, I., Krause, W., & Pook, M. (1999). Cognitive-behavioral therapy for idiopathic infertile couples. Psychotherapy andPsychosomatics, 68, 15-21.

Uchino, B. N., Cacioppo, J. T., & Kiecolt-Glaser, J. K. (1996). Therelationship between social support and physiological processes: Areview with emphasis on underlying mechanisms. Psychological Bulle-tin, 119, 488-531.

Umberson, D. (1992). Gender, marital status and the social control ofhealth behavior. Social Science and Medicine, 24, 907-917.

Unruh, A. M. (1996). Gender variations in clinical pain experience.Pain, 65, 123-167.

Vedhara, K., Cox, N. K. M., Wilcock, G. K., Perks, P., Hunt, M., Ander-son, S., Lightman, S. L., & Shanks, N. M. (1999). Chronic stress inelderly carers of dementia patients and antibody response to influenzavaccination. Lancet, 353, 627-631.

Vitaliano, P. P., Young, H. M., Russo, J., Romano, J., & Magana-Amato,A. (1993). Does expressed emotion in spouses predict subsequent prob-lems among care recipients with Alzheimer's disease? Journal of Ger-ontology, 48, P202-P209.

Waltz, M., Kriegel, W., & Bosch, P. V. (1998). The social environment andhealth in rheumatoid arthritis: Marital quality predicts individual vari-ability in pain severity. Arthritis Care and Research, 11, 356-374.

Ward, M. M., & Leigh, J. P. (1993). Marital status and the progression offunctional disability in patients with rheumatoid arthritis. Arthritis andRheumatism, 36, 581-588.

Weiss, L. H., & Kerns, R. D. (1995). Patterns of pain-relevant socialinteractions. International Journal of Behavioral Medicine, 2, 157-171.

Weissman, M. M. (1987). Advances in psychiatric epidemiology: Ratesand risks for major depression. American Journal of Public Health, 77,445-451.

Wells, K. B., Stewart, A., Hays, R. D., Burnam, A., Rogers, W., Daniels,M., Berry, S., Greenfield, S., & Ware, J. (1989). The functioning andwell-being of depressed patients. Journal of the American MedicalAssociation, 262, 914-919.

Whisman, M. A. (1999). Marital dissatisfaction and psychiatric disorders:Results from the National Comorbidity Survey. Journal of AbnormalPsychology, 108, 701-706.

Wickrama, K., Conger, R. D., & Lorenz, F. O. (1995). Work, marriage,lifestyle, and changes in men's physical health. Journal of BehavioralMedicine, 18, 97-111.

Wickrama, K. A. S., Lorenz, F. O., & Conger, R. D. (1997). Marital qualityand physical illness: A latent growth curve analysis. Journal of Marriageand the Family, 59, 143-155.

Williamson, D., Robinson, M. E., & Melamed, B. (1997). Pain behavior,spouse responsiveness, and marital satisfaction in patients with rheuma-toid arthritis. Behavioral Modification, 21, 97-118.

Zautra, A. J., Burleson, M. H., Matt, K. S., Roth, S., & Burrows, L. (1994).Interpersonal stress, depression, and disease activity in rheumatoid ar-thritis and osteoarthritis patients. Health Psychology, 13, 139-148.

Zautra, A. J., Hoffman, J. M., Matt, K. S., Yocum, D., Potter, P. T., Castro,W. L., & Roth, S. (1998). An examination of individual differences inthe relationship between interpersonal stress and disease activity amongwomen with rheumatoid arthritis. Arthritis Care and Research, 11,271-279.

Received August 19, 1999Revision received November 7, 2000

Accepted November 27, 2000 •

marriage and health: his and hers

Documents